A class of nonlinear multi-objective mixed integer programming model for oil and gas portfolio and algorithm solution

Abstract

This paper is concerned with the oil & gas assets portfolio. A multi-objective portfolio model of oil & gas assets is studied from two perspectives—scale and revenue. Considering the nonlinear and integer constraints in the model, a class of oil & gas assets portfolio model of nonlinear multi-objective mixed integer programming is established. The weight of the multi-objective is solved by the support vector machine model. A hybrid genetic algorithm, which uses the position displacement strategy of the particle swarm optimizer as a mutation operation, is applied to the optimization model. Finally, two examples are applied to verify the effectiveness of the model and algorithm.

Keywords

Oil and gas portfolio support vector machine nonlinear multi-objective mixed integer programming

Introduction

One of the common questions in the oil & gas portfolio is how to allocate some amount of money to different assets. There are many different models and methodologies for the analysis of the oil and gas portfolio. The first approach for considering the optimal portfolio problem is the so-called mean-variance. It was pioneered by Markowitz (1952) and he played a critical role in the theory of Portfolio selection.

In the Markowitz portfolio model, the risk dispersion problem with the upstream field of the oil & gas industry is simulated and decided. The optimal portfolio is consistent with the minimum risk under the determined return or the maximum return under the determined risk (Guo et al., 2012; Markowitz, 1952).

Conditional value at risk (CVaR) is the most frequently used risk measure in current risk management practice. It evaluates the average loss comprehensively when the potential decline of oil & gas assets are higher than the value at risk values under a given confidence level during a certain investment period. Then the portfolio can be re-adjusted asset ratios based on this loss (Li, 2013; Wu et al., 2018; Yang, 2017).

The Boston Consulting Group (BCG) Matrix divides different strategic business units into four categories and presents them intuitively. It can help strategy makers analyze the market performance and cash flow of products of their portfolio through the matrix (Li et al., 2019; Zhang, 2007).

The General Electric Matrix is a comprehensive portfolio planning tool jointly developed by McKinsey and General Electric in the early 1970s. It primarily uses the rate of return on investment as the norm for assessing investment opportunities. It adds more evaluation indicators to the BCG growth-share model, adopts comprehensive measurement standards of the classification of business units, and introduces the middle position to make the evaluation more accurate (Zhang , 2007).

The Arthur D. Little (ADL) Matrix was proposed by Arthur D. International Management Consulting Company in the 1970s. The model relies on the industry life cycle for portfolio management. The structure of the autoregressive distributed lag model includes two main analysis indexes. One is an environmental indicator that represents the industry life cycle. Another is the competitive position indicator which reflects the strength of enterprises (Zhang , 2007).

The Shell Directional Policy Matrix (DPM) is proposed by Royal Dutch Shell in the 1970s. It makes a qualitative analysis of the business based on the business prospect attraction and company competitiveness. Each dimension of the DPM is divided into three categories. Enterprises can evaluate the characteristics of nine types of businesses in the matrix and formulate the corresponding policies strategies and suggestions regarding the direction provided by the matrix (Tuncay, 2013; Zhang , 2007).

The above portfolio optimization methodologies have different purposes. They focus on the internal benefit evaluation under the consideration of the external market, the quality evaluation of assets themselves, and the income evaluation under the consideration of risk. Compared with the return of optimization purposes. The above portfolio optimization methodologies are more concentrated and the quantitative consideration of the multi-objective of the oil & gas portfolio is insufficient. There is a lack of multi-objective decision-making. In some assets, the decision variables are often limited to invest or not, which the mixed 0–1 integer decision should be considered in the portfolio (Bigerna et al, 2021; Gazijahani et al., 2017; Globocnik et al., 2020; Kiamehr et al, 2018; Ning et al., 2020; Yan, 2018).

In this paper, a class of nonlinear multi-objective mixed integer programming models for oil & gas portfolios is formulated. The outline of this paper is as follows. Then, we formulate a mathematical model to describe the oil & gas portfolio framework, including the nonlinear multi-objective and mixed integer decision-making. Nest, this paper devotes a support vector machine (SVM) model to obtain the weights of the multi-objective. Moreover, the hybrid genetic algorithm (GA) with particle swarm optimizer (PSO) is applied to the portfolio model. Two numerical experiments are studied. It concludes the whole paper at last.

A mathematical model

An oil company will make a portfolio decision on $n$ assets (Rijnsburger, 2017; Yu et al, 2021). Suppose that $x_{i} (i = 1, \dots, n)$ denotes the equity ratio of asset $i$ . The corresponding oil production, gas production, oil reserves, gas reserves, net present value (NPV) of the remaining contract period, and accumulative net cash flow of the remaining contract period are $P O_{i}$ , $P G_{i}$ , $R O_{i}$ , $R G_{i}$ , $N P V_{i}$ , and $A N C F_{i}$ , respectively. Because their units are not unified, it needs to be normalized according to the following equation:

X_{i} = \frac{N_{i} - min N_{i}}{max N_{i} - min N_{i}}

(1)

where

X_{i}

represents the normalized data of the index,

N_{i}

represents the original data of the index,

max

represents the maximum value, and

min

represents the minimum value. Equation (1) can unify the multiple repetitions and miscellaneous of the original data. It can provide a concise and unified data source for the next modeling and solution.

After normalization, oil production, natural gas production, oil reserves, natural gas reserves, NPV of the remaining contract period, and cumulative net cash flow of the remaining contract period are $S T P O_{i}, S T P G_{i}, S T R O_{i}, S T R G_{i}, S T N P V_{i}$ , and $S T A N C F_{i}$ , respectively. The corresponding equations are as follows:

{\begin{aligned} S T P O_{i} & = \frac{P O_{i} - min P O_{i}}{max P O_{i} - min P O_{i}} \cdot 10 \\ S T P G_{i} & = \frac{P G_{i} - min P G_{i}}{max P G_{i} - min P G_{i}} \cdot 10 \\ S T R O_{i} & = \frac{R O_{i} - min R O_{i}}{max R O_{i} - min R O_{i}} \cdot 10 \\ S T R G_{i} & = \frac{R G_{i} - min R G_{i}}{max R G_{i} - min R G_{i}} \cdot 10 \\ S T N P V_{i} & = \frac{N P V_{i} - min N P V_{i}}{max N P V_{i} - min N P V_{i}} \cdot 10 \\ S T A N C F_{i} & = \frac{A N C F_{i} - min A N C F_{i}}{max A N C F_{i} - min A N C F_{i}} \cdot 10 \end{aligned}

(2)

Oil production, natural gas production, oil reserves, and natural gas reserves are the scale objectives. NPV of the remaining contract period and cumulative net cash flow of the remaining contract period are the revenue objectives. The objective function is to maximize the scale and revenue of the portfolio. The scale of the portfolio includes 4 objectives and the return of the portfolio includes 2 objectives, as shown below:

\begin{aligned} F_{S} = & ω_{1} \cdot F_{S 1} + ω_{2} \cdot F_{S 2} + ω_{3} \cdot F_{S 3} + ω_{4} \cdot F_{S 4} = ω_{1} \cdot \sum_{i = 1}^{n} (S T P O_{i} \cdot x_{i}) \\ + ω_{2} \cdot \sum_{i = 1}^{n} (S T P G_{i} \cdot x_{i}) + ω_{3} \cdot \sum_{i = 1}^{n} (S T R O_{i} \cdot x_{i}) + ω_{4} \cdot \sum_{i = 1}^{n} (S T R G_{i} \cdot x_{i}) \end{aligned}

(3)

F_{B} = ω_{5} \cdot F_{B 1} + ω_{6} \cdot F_{B 2} = ω_{5} \cdot \sum_{i = 1}^{n} (S T N P V_{i} \cdot x_{i}) + ω_{6} \cdot \sum_{i = 1}^{n} (S T A N C F_{i} \cdot x_{i})

(4)

where

(ω_{1}, ω_{2}, ω_{3}, ω_{4}, ω_{5}, ω_{6})

is the weight of the objective and

\sum_{i = 1}^{6} ω_{i} = 1

is satisfied. Considering oil production, natural gas production, oil reserves, natural gas reserves, NPV of the remaining contract period, and accumulative net cash flow in the remaining contract period, the following relationship should be satisfied:

{\begin{aligned} \frac{P O_{i}}{x_{i}} & = \frac{I N I P O_{i}}{I N I X_{i}} \\ \frac{P G_{i}}{x_{i}} & = \frac{I N I P G_{i}}{I N I X_{i}} \\ \frac{R O_{i}}{x_{i}} & = \frac{I N I R O_{i}}{I N I X_{i}} \\ \frac{R G_{i}}{x_{i}} & = \frac{I N I R G_{i}}{I N I X_{i}} \\ \frac{N P V_{i}}{x_{i}} & = \frac{I N I N P V_{i}}{I N I X_{i}} \\ \frac{A N C F_{i}}{x_{i}} & = \frac{I N I A N C F_{i}}{I N I X_{i}} \end{aligned}

(5)

where

I N I P O_{i}, I N I P G_{i}, I N I R O_{i}, I N I R G_{i}, I N I N P V_{i}

, and

I N I A N C F_{i}

represent the initial value of oil production, natural gas production, oil reserves, natural gas reserves, NPV of the remaining contract period, and accumulative net cash flow in the remaining contract period, respectively,

I N I X_{i}

represents the initial equity ratio of asset

i

The model sets some constraints, including the ratio of high-risk assets, the ratio of oil assets, the ratio of conventional assets, the ratio of exploration assets, the ratio of onshore assets, the scope of equity ratio after optimization, the scope of equity ratio change after optimization, and the scope of equity ratio change for some specific assets and so on.

For the ratio of high-risk assets, it is satisfied as follows by setting an interval $[R_{L O W}, R_{H I G H}]$ :

{\begin{matrix} \frac{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i} \cdot I R_{i}}{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i}} \in [R_{L O W}, R_{H I G H}] \\ I R_{i} = {\begin{matrix} 1, asset i is high - risk \\ 0, asset i is not high - risk \end{matrix} \end{matrix}

(6)

For the ratio of oil assets, it is satisfied as follows by setting an interval $[O_{L O W}, O_{H I G H}]$ :

\begin{aligned} {\begin{matrix} \frac{\sum_{i = 1}^{n} P O_{i} \cdot x_{i} \cdot I O_{i}}{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i}} \in [O_{L O W}, O_{H I G H}] \\ I O_{i} = {\begin{cases} 1, asset i is high - risk \\ 0, asset i is not high - risk \end{cases} \end{matrix} \end{aligned}

(7)

For the ratio of conventional assets, it is satisfied as follows by setting an interval $[C_{L O W}, C_{H I G H}]$ :

\begin{aligned} {\begin{matrix} \frac{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i} \cdot I C_{i}}{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i}} \in [C_{L O W}, C_{H I G H}] \\ I C_{i} = {\begin{cases} 1, asset i is high - risk \\ 0, asset i is not high - risk \end{cases} \end{matrix} \end{aligned}

(8)

For the ratio of exploration assets, it is satisfied as follows by setting an interval $[R E_{L O W}, R E_{H I G H}]$ :

\begin{aligned} {\begin{matrix} \frac{\sum_{i = 1}^{n} (R O_{i} + R G_{i}) \cdot x_{i} \cdot I R E_{i}}{\sum_{i = 1}^{n} (R O_{i} + R G_{i}) \cdot x_{i}} \in [R E_{L O W}, R E_{H I G H}] \\ I R E_{i} = {\begin{cases} 1, asset i is high - risk \\ 0, asset i is not high - risk \end{cases} \end{matrix} \end{aligned}

(9)

For the ratio of onshore assets, it is satisfied as follows by setting an interval $[M_{L O W}, M_{H I G H}]$ :

\begin{aligned} {\begin{matrix} \frac{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i} \cdot I M_{i}}{\sum_{i = 1}^{n} (P O_{i} + P G_{i}) \cdot x_{i}} \in [M_{L O W}, M_{H I G H}] \\ I M_{i} = {\begin{cases} 1, asset i is high - risk \\ 0, asset i is not high - risk \end{cases} \end{matrix} \end{aligned}

(10)

For the scope of equity ratio after optimization, it is satisfied as follows by setting an interval $[I N I_{i, L O W}, I N I_{i, H I G H}]$ :

I N I_{i, L O W} \leq x_{i} \leq I N I_{i, H I G H}, i \in [1, 2, \dots, n]

(11)

For the equity ratio change after optimization, it is satisfied as follows by setting an interval $[O P T_{i, L O W}, O P T_{i, H I G H}]$ :

O P T_{i, L O W} \leq x_{i} \leq O P T_{i, H I G H}, i \in [1, 2, \dots, n]

(12)

For the scope of equity ratio change about some specific assets, the decision variable is limited to investment or not. Therefore, the equity ratio is considered to be $0$ or $1$ under some special decision-making goal. Then $x_{i}$ should be satisfied by

x_{i} (x_{i} - 1) = 0, i \in [1, 2, \dots, n]

(13)

In summary, we can establish an optimization model of the portfolio as follows:

\begin{aligned} \max (F_{S}, F_{B}) & = \max (ω_{1} \cdot F_{S 1} + ω_{2} \cdot F_{S 2} + ω_{3} \cdot F_{S 3} + ω_{4} \cdot F_{S 4} + ω_{5} \cdot F_{B 1} + ω_{6} \cdot F_{B 2}), \\ = \max [ω_{1} \cdot \sum_{i = 1}^{n} (S T P O_{i} \cdot x_{i}) + ω_{2} \cdot \sum_{i = 1}^{n} (S T P G_{i} \cdot x_{i}) + ω_{3} \cdot \sum_{i = 1}^{n} (S T R O_{i} \cdot x_{i}) \\ + ω_{4} \cdot \sum_{i = 1}^{n} (S T R G_{i} \cdot x_{i}) + ω_{5} \cdot \sum_{i = 1}^{n} (S T N P V_{i} \cdot x_{i}) + ω_{6} \cdot \sum_{i = 1}^{n} (S T A N C F_{i} \cdot x_{i})] \end{aligned}

s . t . {\begin{matrix} S T P O_{i} = \frac{P O_{i} - min P O_{i}}{max P O_{i} - min P O_{i}} \cdot 10 \\ S T P G_{i} = \frac{P G_{i} - min P G_{i}}{max P G_{i} - min P G_{i}} \cdot 10 \\ S T R O_{i} = \frac{R O_{i} - min R O_{i}}{max R O_{i} - min R O_{i}} \cdot 10 \\ S T R G_{i} = \frac{R G_{i} - min R G_{i}}{max R G_{i} - min R G_{i}} \cdot 10 \\ S T N P V_{i} = \frac{N P V_{i} - min N P V_{i}}{max N P V_{i} - min N P V_{i}} \cdot 10 \\ S T A N C F_{i} = \frac{A N C F_{i} - min A N C F_{i}}{max A N C F_{i} - min A N C F_{i}} \cdot 10 \\ \frac{P O_{i}}{x_{i}} = \frac{I N I P O_{i}}{I N I X_{i}} \\ \frac{P G_{i}}{x_{i}} = \frac{I N I P G_{i}}{I N I X_{i}} \\ \frac{R O_{i}}{x_{i}} = \frac{I N I R O_{i}}{I N I X_{i}} \\ \frac{R G_{i}}{x_{i}} = \frac{I N I R G_{i}}{I N I X_{i}} \\ \frac{N P V_{i}}{x_{i}} = \frac{I N I N P V_{i}}{I N I X_{i}} \\ \frac{A N C F_{i}}{x_{i}} = \frac{I N I A N C F_{i}}{I N I X_{i}} \\ 0 \leq x_{i} \leq 1, i = 1, 2, \dots, n, \\ x_{i} statisfy (7) - (13), i \in [1, 2, \dots, n] . \end{matrix}

(14)

A calculation methodology

It is found that model (14) is a highly nonlinear optimization model, which can not be solved by a gradient algorithm. In addition, it is found that the weight proportion $(ω_{1}, ω_{2}, ω_{3}, ω_{4}, ω_{5}, ω_{6})$ in the objective function also needs to be solved.

A solution of the objective function weight

There are two ways to obtain the weight of the objective function $(ω_{1}, ω_{2}, ω_{3}, ω_{4}, ω_{5}, ω_{6})$ . One is to determine the weight through the experts’ system. The other is solved by a certain mathematical model. Here, we use a SVM model (14) (Cristianini et al., 2000; Fan et al., 2005; Fayad, 1999).

First, we suppose that H-asset is the asset with both high scale value and high return, L-asset is the asset with both low scale value and low return. For $n$ assets, we can select some extremely H- and L-assets through an expert system. Although different experts have different opinions, some extremely H- and L-assets are well acknowledged universally by all experts. We screen out some extremely H- and L-assets, and their sets are $n_{1}$ and $n_{2}$ , respectively. We set the following values:

{\begin{matrix} y_{i} = 1, i \in n_{1} \\ y_{i} = - 1, i \in n_{2} \end{matrix}

Suppose that the scale and revenue vector of an asset is

z

. As shown in Figure 1, the H-assets and L-assets can be divided linearly (if linear separation is not possible, it can be divided linearly by the appropriate mapping of H- and L-assets). The solid and hollow circles represent H-assets and L-assets, respectively. The plane

ω z + b = 0

is the critical plane for H- and L-assets, where

ω z

denotes the inner product of

ω

and

z

. The farther a point is from the plane, the more difficult it is for it to cross the plane (i.e. it is difficult to transfer an H-asset from L-asset, and vice versa). Meanwhile, these points represent the extreme H-assets from L-assets. The other assets should be closer to the plane because they have no significant H- or L-characteristics.

Figure 1.

Schematic diagram of the risk assessment using the support vector machine model.

We can prove the following Theorems 1 to 4 easily. Theorem 1 shows that the countries that belong to the same scale and return category are on the same side of the critical plane. Moreover, the assets that belong to different scale and return categories are on different sides of the critical plane. Theorems 2 to 4 provide the theoretical foundation of the SVM.

Theorem 1

For an H-asset $x_{H}$ and an L-asset $z_{L}$ , H- and L-assets are linearly separable. The critical plane is $ω z + b = 0$ . The definition of function $f (x)$ is shown as $ω z + b$ , where

f (x_{H}) > 0, f (z_{L}) < 0

(15)

As mentioned earlier, for both H- and L-assets, the farther from the plane $ω z + b = 0$ the point is, the more difficult it is to cross this plane, as indicated by Theorems 2 and 3.

Theorem 2

For two H-assets A and B, the scale and return vector are $z_{H A}$ and $z_{H B}$ , respectively. The critical plane is $ω z + b = 0$ . By defining the function $f (x)$ is shown as $ω z + b$ , when the value of associated scale and return with A is higher than B, we have

| f (z_{H A}) | > | f (z_{H B}) |

(16)

Theorem 3

For two L-assets A and B, the scale and return vector are $z_{L A}$ and $z_{L B}$ , respectively. The critical plane is $ω z + b = 0$ . By defining the function $f (x)$ is shown as $ω z + b$ , when the value of associated scale and return with A is lower than B, we have

| f (z_{L A}) | > | f (z_{L B}) |

(17)

It is not difficult to prove Theorems 2 and 3, based on the distance between the H-/L-assets and the critical plane $z + b = 0$ . From Theorems 1 to 3, we can derive Theorem 4 as follows:

Theorem 4

For two assets A and B, the scale and return vector are $z_{A}$ and $z_{B}$ , respectively. The critical plane is $ω z + b = 0$ . By defining the function $f (x)$ is shown as $ω z + b$ , when the value of associated scale and return with A is higher than B, we have

f (z_{A}) > f (z_{B})

(18)

From Theorem 4, we initially obtained the critical plane $ω z + b = 0$ , concerning H- and L-assets. Subsequently, the key is achieving the critical plane $ω z + b = 0$ accurately. The following part focuses on solving the critical plane.

As shown in Figure 1, two support planes $ω z + b = 1$ and $ω z + b = - 1$ can be found on either side of the critical plane $ω z + b = 0$ . The distance between the two support planes is $2 / ‖ ω ‖$ . The optimal critical plane should be the one in the middle, which creates the longest distance between the support planes. Therefore, in accordance with the idea of maximizing the “interval,” the following optimization model can be established:

\begin{aligned} max_{ω, b} \frac{2}{‖ ω ‖} \\ s . t . {\begin{matrix} for a H - asset i, ω^{T} z_{i} + b \geq 1 \\ for a L - asset i, ω^{T} z_{i} + b \geq 1 \end{matrix} \end{aligned}

(19)

The optimization problem is equivalent to the following:

\begin{aligned} min_{ω, b} \frac{‖ ω ‖}{2} \\ s . t . {\begin{matrix} y_{i} (ω^{T} z_{i} + b) \geq 1, i = 1, \dots, m \end{matrix} \end{aligned}

(20)

where

y_{i} = 1

and

y_{i} = - 1

indicate that the

i

th asset is an H-asset and L-asset, respectively.

m

is the number of assets in the training set.

Model (20) uses convex quadratic programming. However, we selected a hyperplane to divide the training set accurately and completely. However, in practical applications, there may not be such a perfect situation, which means it could be a linear non-separable problem. In other words, some hollow circles might be in the solid-circle region and vice versa. In that case, some training points that do not satisfy the constraints are allowed to exist. Therefore, we introduced slack variables to identify the weakening constraint condition (Platt, 1998; Xu et al., 2021; Zhang, 2021).

y_{i} (ω^{T} z_{i} + b) \geq 1, i = 1, \dots, m

(21)

It is apparent that the training point

(z_{i}, y_{i})

can satisfy the constraint condition (21). However, the ”wrong points” degree should be avoided, ensuring

ξ_{i}

is not too large. The objective function is added to a penalty function

\sum_{i = 1}^{m} ξ_{i}

. Therefore, Model (??) can be rewritten as an optimization model as follows:

\begin{aligned} min_{ω, b, ξ} \frac{1}{2} ω^{T} ω + C \sum_{i = 1}^{m} ξ_{i} \\ s . t . {\begin{matrix} y_{i} (ω^{T} z_{i} + b) \geq 1 - ξ_{i}, i = 1, \dots, m \\ ξ_{i} \geq 0, i = 1, \dots, m \end{matrix} \end{aligned}

(22)

where the constant

C

is a penalty parameter. Model (22) wants to minimize

\frac{1}{2} ω^{T} ω

(maximize interval

\frac{2}{‖ ω ‖}

) and

\sum_{i = 1}^{m} ξ_{i}

(minimize the damage degree of constraints

y_{i} (ω^{T} z_{i} + b) \geq 1

). The penalty parameter

C

reflects our trade-off between them. However, it is difficult to solve the above-mentioned problem. We can attempt to solve using the dual model by introducing a Lagrange function as follows:

\begin{aligned} L (ω, b, ξ, α, β) = \frac{1}{2} ω^{T} ω + C \sum_{i = 1}^{m} ξ_{i} - \sum_{i = 1}^{m} α_{i} [y_{i} (ω^{T} z_{i} \\ + b) - 1] - Σ_{i = 1}^{m} β_{i} ξ_{i}, i = 1, \dots, m \end{aligned}

(23)

where

α = {α_{i}}

β = {β_{i}}

. The dual problem of the original problem is

\begin{aligned} max_{α, β} min_{ω, b, ξ} L (ω, b, ξ, α, β) \\ s . t . α_{i} \geq 0, β_{i} \geq 0 \end{aligned}

(24)

The partial derivation can be obtained as follows:

\begin{aligned} \frac{\partial L}{\partial ω = ω - \sum_{i = 1}^{m} α_{i} y_{i} z_{i} = 0 \Rightarrow ω = \sum_{i = 1}^{m} α_{i} y_{i} z_{i} \frac{\partial L}{\partial b} = - \sum_{i = 1}^{m} α_{i} y_{i} = 0 \Rightarrow \sum_{i = 1}^{m} α_{i} y_{i} = 0 \frac{\partial L}{\partial ξ}} \\ = C - α_{i} - β_{i} = 0 \Rightarrow C = α_{i} + β_{i} r \end{aligned}

(25)

Substituting (25) into (23), we can obtain

L (ω, b, ξ, α, β) = - \frac{1}{2} \sum_{i = 1}^{m} \sum_{i = 1}^{m} α_{i} α_{j} y_{i} y_{j} z_{i}^{T} z_{j} + \sum_{i = 1}^{m} α_{i}

(26)

Then the dual model is as follows:

\begin{aligned} min_{α} \frac{1}{2} \sum_{i = 1}^{m} \sum_{i = 1}^{m} α_{i} α_{j} y_{i} y_{j} z_{i}^{T} z_{j} - \sum_{i = 1}^{m} α_{i} \\ s . t . {\begin{matrix} \sum_{i = 1}^{m} α_{i} y_{i} = 0 \\ 0 \leq α_{i} \leq C, i = 1, \dots, m \end{matrix} \end{aligned}

(26)

Model (26) can be rewritten as

\begin{aligned} min_{α} \frac{1}{2} α^{T} Q α - e^{T} α \\ s . t . {\begin{matrix} y^{T} α = 0 \\ 0 \leq α_{i} \leq C, i = 1, \dots, m \end{matrix} \end{aligned}

(27)

where

Q_{i j} = y_{i} y_{j} z_{i}^{T} z_{j}

e = [1, \dots, 1]

Thus, the original problem is transformed into a relatively simple quadratic programming model. Therefore, we initially solved a programming model (27) to obtain $α$ . Subsequently, according to formula (25), we can obtain $ω$ .

The GA with PSO

The GA originated in the 1960s and is a global random search heuristic algorithm, which draws on the idea of evolution. It is a search technique used in computing to find exact or approximate solutions to optimization and search problems. GAs are categorized as global search heuristics. GAs are a particular class of evolutionary algorithms that use techniques inspired by evolutionary biology such as inheritance, mutation, selection, and crossover. The GA is implemented in a computer simulation in which a population of abstract representations of candidate solutions to an optimization problem evolves toward better solutions. However, the GA has some inherent flaws, such as slow convergence. We draw the displacement of the particle swarm algorithm to improve the GA to modify the thinking of mutation.

The PSO is a population-based algorithm. It was studied by Kennedy et al. (1995). PSO algorithm relies on a group of particles to modify the displacement continuously in the solution space according to the group experience and the individual optimization process, to realize the search for the optimal solution in multi-dimensional space. The swarm is initialized firstly in a set of randomly generated potential solutions and each particle in the population is a feasible solution to the optimization problem. Then a fitness function is used to calibrate the fitness value of the particle to judge whether the particle is optimal. Each particle will move in the solution space and correct its displacement constantly until the optimal solution is found. The modification of displacement depends on a velocity vector, which is determined by the optimal solution found so far by the particle itself and the optimal solution found so far by the whole population. The former represents the cognitive level of individual particles and the latter represents the cognitive ability of groups.

Suppose that the population size is $N$ and the particle dimension is $K$ . The position of particle $i$ th is

P O S_{i} = [p o s_{i}^{1}, p o s_{i}^{2}, \dots, p o s_{i}^{K}]

and its velocity vector is

S_{i} = [s_{i}^{1}, s_{i}^{2}, \dots, s_{i}^{K}]

The best position for each individual is

p O p t_{i} = [p O p t_{i}^{1}, p O p t_{i}^{2}, \dots, p O p t_{i}^{K}]

The best solution for the swarm is

g O p t_{i} = [g O p t_{i}^{1}, g O p t_{i}^{2}, \dots, g O p t_{i}^{K}]

In the PSO, the particles

S_{i}

are manipulated according to the following equation:

{\begin{aligned} S_{i} & = w e i g h t \times S_{i} + A C_{1} \times R A_{1} \times (p O p t_{i} - P O S_{i}) + A C_{2} \times R A_{2} \times (g O p t_{i} - P O S_{i}) \\ P O S_{i} & = P O S_{i} + S_{i} c a s e s \end{aligned}

(28)

where

A C_{1}

and

A C_{2}

are two positive constants called acceleration factors, weight is the inertia weight, and

R A_{1}

and

R A_{2}

are uniformly distributed numbers in

[0, 1]

. In addition, a maximum velocity

S_{max}

is introduced to limit the particle velocity

S_{i}

. If the current acceleration of a particle causes the speed of one of its components to exceed the maximum speed of that component, the speed of that component will be set to the maximum speed of that component.

Krohling (2004) proposed a new position displacement principle. He suggested that the updating of the velocity vector is independent of the history velocity vector. According to his work, the particles are manipulated as follows:

{\begin{aligned} S_{i} & = | R A_{1} | \times (p O p t_{i} - P O S_{i}) + | R A_{2} | \times (g O p t_{i} - P O S_{i}) \\ P O S_{i} & = P O S_{i} + S_{i} c a s e s \end{aligned}

(29)

where

| R A_{1} |

and

| R A_{2} |

are positive random numbers generated according to the absolute value of the Gaussian probability distribution GAUSS(0,1).

It can be seen that the optimization of particles depends on position transfer from the standard PSO algorithm. Angeline (1998a,b) pointed out that the position displacement of particles was a kind of mutation operation. We make use of the position displacement as a mutation operator to modify the GA. During the mutation operation, the position of those individuals selected for mutation is modified via a vector that depends on the personal best position and the current global best position. If GA adopts the new mutation operator, it can use the history information of the chromosomes and share information among the population. The mutation will no longer be stochastic and the local search will be more efficient. At the same time, it keeps the GA’s good property in global search. So the hybrid GA can provide more efficient exploration and exploitation ability (Ding, 2018; Wei et al., 2022).

However, the velocity vector has to be treated specially during the mutation operation. After selection and crossover, the chromosomes are different from their parents. The history velocity vectors present the parents’ information and will not make sense anymore. So during the mutation, equation (29) is not suitable. But equation (30) is available for the operation. The hybrid GA with PSO is described as follows:

Step 1. Initialize the population in the search space.

Step 2. Record every individual’s best solution $p O p t_{i}$ and the best solution among the population $g O p t_{i}$ .

Step 3. Perform selection operation.

Step 4. According to the crossover probability pc, perform crossover operation.

Step 5. Select individuals for mutation and employ the results coming from Step 2 to perform the mutation operation according to equation (30).

Step 6. If the optimum is found, quit the iterative operation; if not, return to Step 2.

Numerical experiments

To verify the effectiveness of the nonlinear multi-objective mixed integer programming model for oil & gas portfolio, the algorithm for objective function weight with support vector machine (SVM) and the hybrid GA with PSO discussed in the previous section, two numerical experiments applied carried out in this section. All the computations are performed in Matlab on a Pentium PC under the Windows 10 environment.

S company oil & gas portfolio

Firstly, we take the international oil company-S as an example to optimize 170 assets from 25 countries in the oil & gas upstream. The data of 170 assets are obtained from IHS Markit, World Bank, and Wood Mackenzie. For these 170 assets, we select 10 extremely H-assets and L-assets and substitute them into the model (27) to solve the corresponding problems $α$ . Then the corresponding weight value can be obtained according to equation (25).

Then, we solve model (14) and set the constraint interval for the ratio of high-risk assets as $[0, 0.25]$ , the ratio of oil assets as $[0.5, 0.8]$ , and the ratio of conventional oil & gas assets as $[0.45, 0.7]$ . In this case study, it is assumed that the equity ratio of 9 assets is $1$ or $0$ . After applying the calculation method in the section entitled ’The GA with PSO’ to optimize the equity ratio in 170 assets, the changes in the scale and return are shown in Tables 1 and 2:

Table 1.

The weight of scale and revenue of S company.

Scale				Revenue
Oil Production ( $ω_{1}$ )	Gas production ( $ω_{2}$ )	Oil reserves ( $ω_{3}$ )	Gas reserves ( $ω_{4}$ )	Net present value (NPV) of the remaining contract period ( $ω_{5}$ )	Accumulative net cash flow in the remaining contract period ( $ω_{6}$ )
0.0733	0.1051	0.1835	0.2051	0.2068	0.2262

Table 2.

The changes in scale and revenue of S company.

Type	Item	Before optimization	After optimization	Changes
Scale	Oil production (kboe/d)	1804	2178	20.7%
Scale	Gas production (kboe/d)	1612	2119	31.5%
	Oil reserves (mmboe)	15,392	20,792	35.1%
	Gas reserves (mmboe)	20,602	25,136	22.0%
Revenue	Net present value (million dollars)	143,132	167112	16.8%
	Accumulative net cash flow (million dollars)	277,564	318,285	14.7%

kboe/d is kilo barrels of oil equivalent per day, mmboe is million barrels of oil equivalent.

C company oil & gas portfolio

Next, we take a national oil company-C as another example to optimize 42 assets from 20 countries in the oil & gas upstream. The data of 42 assets are also obtained from IHS Markit, World Bank, and Wood Mackenzie. For these 42 assets, we select 6 extremely H-assets and L-assets and substitute them into model (27) to solve the corresponding problems $α$ . Then the corresponding weight value can be obtained as following Table 3.

Table 3.

The weight of scale and revenue of C company.

Scale				Revenue
Oil production ( $ω_{1}$ )	Gas production ( $ω_{2}$ )	Oil reserves ( $ω_{3}$ )	Gas reserves ( $ω_{4}$ )	Net present value (NPV) of the remaining contract period ( $ω_{5}$ )	Accumulative net cash flow in the remaining contract period ( $ω_{6}$ )
0.0966	0.1121	0.1625	0.2209	0.1992	0.2087

By setting the constraint interval for the ratio of high-risk assets as $[0, 0.8]$ , the ratio of oil assets as $[0.3, 0.9]$ , and the ratio of conventional oil & gas assets as $[0.3, 1]$ , we can solve model (14). In this case study, it is also assumed that the equity ratio of 9 assets is $1$ or $0$ . The changes in the scale and revenue are shown in Table 4.

Table 4.

The changes in scale and revenue of C company.

Type	Item	Before optimization	After optimization	Changes
Scale	Oil production (kboe/d)	5940	6904	16.2%
Scale	Gas production (kboe/d)	1295	1338	3.3%
	Oil reserves (mmboe)	18,958	19,451	2.6%
	Gas reserves (mmboe)	5335	9284	74%
Revenue	Net present value (million dollars)	21,186	21,656	2.2%
	Accumulative net cash flow (million dollars)	45,827	48,738	6.4%

It can be seen from the above table that the scale and revenue of S company and C company have been optimized after applying the models and algorithms in this paper. In the case of S company, the oil & gas production has been improved greatly. However, in the case of C company, the oil production has a significant increase after optimization while the increase in gas production is not significant. The increase in oil reserves is not significant while there has been a significant increase in gas reserves.

Conclusion

This paper establishes an optimization model based on project level for the investment decision of upstream oil & gas of international oil companies. A nonlinear multi-objective mixed integer programming model with scale and revenue as objectives and structure as constraints are established. The method of SVM is used for the weight of the target, and the hybrid GA based on the idea of particle swarm displacement transfer is used for the planning model to construct the calculation system. The case of an oil company is verified. From the results, our method is effective and feasible.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Wei Yan

References

Angeline

(1998) Evolutionary optimization versus particle swarm optimization: philosophy and performance differences. Evolutionary Programming 7: 601–610.

Angeline

(1998) Using selection to improve particle Swarm optimization. In: Proceedings of IEEE international conference on evolutionary compotation, May 04-09, 1998, Anchorage, AK, USA, 84-89. New York, NY: IEEE.

Bigerna

Bollino

Polinori

(2021) Oil import portfolio risk and spillover volatility. Resources Policy 70: 1–11.

Cristianini

Shawe-Taylor

(2000) An Introduction to Support Vector Machines. London: Cambridge University Press.

Ding

Schulz

Shen

, et al. (2018) Energy aware scheduling in flexible flow shops with hybrid particle swarm optimization. Computers and Operations Research 125: 1–17.

Fan

Chen

Lin

(2005) Working set selection using second order information for training SVM. Journal of Machine Learing Research 12: 1889–1918.

Fayad

Schmidt

Johnson

(1999) Building Application Frameworks: Object-Oriented Foundations of Framework Design. New York: John Wiley and Sons.

Gazijahani

Ravadanegh

Salehi

(2017) Stochastic multi-objective model for optimal energy exchange optimization of networked microgrids with presence of renewable generation under risk-based strategies. ISA Transactions 73: 100–111.

Globocnik

Faullant

Parastuty

(2020) Bridging strategic planning and business model management—a formal control framework to manage business model portfolios and dynamics. European Management Journal 38: 231–243.

10.

Guo

Zou

, et al. (2012) A class of multi-period semi-variance portfolio for petroleum exploration and development. International Journal of Systems Science 43: 1883–1890.

11.

(2013) The empirical study of electricity market’s financial risk assessment based on CVaR. Advanced Materials Research 805-806: 1103–1106.

12.

Wang

, et al. (2019) Classified market development strategy of international oil and gas engineering construction equity investment projects based on Boston matrix. Journal of Oil and Gas Technology 41: 38–41.

13.

Kennedy

Eberhart

(1995) Particle swarm optimization. In: Proc. IEEE Conf. on Neural Networks, Perth, November 27-December 01, 1995, WA, Australia, 1942–1948. New York, NY: IEEE.

14.

Kiamehr

Azar

Nayeri

(2018) A multi-objective optimization model for sesigning business portfolio in the oil industry. Engineering, Technology and Applied Science Research 8: 3657–3667.

15.

Krohling

(2004) Gaussian swarm: A novel particle swarm optimization Algorithm. In: Proceedings of the 2004 IEEE conference on Cybernetics and Intelligent Systems, December 01-03, 2004, Singapore, 372-376. New York, NY: IEEE.

16.

Markowitz

(1952) Portfolio selection. Journal of Finance 7: 77–91.

17.

Ning

Chen

Zhang

, et al. (2020) Energy conservation and emission reduction path selection in China: A simulation based on bi-level multi-objective optimization model. Energy policy 137: 111–116.

18.

Platt

(1998) Fast Training of Support Vector Machines Using Sequential Minimal Optimization. Cambridge, MA: MIT Press.

19.

Rijnsburger

(2017) Strategic portfolio management of upstream oil companies; Based on a carbon-constrained world and stranded oil reserves. Business 2017: 1–7.

20.

Tuncay

(2013) A modified strategic position and action evaluation (SPACE) matrix method. In: Proceedings of the international multi conference of engineers and computer scientists, Vol II, IMECS 2013, March 13-15, 2013, HongKong, 866–869. Hong Kong: International Association of Engineers.

21.

Wei

Cao

(2022) Influencing factors analysis and forecasting of residential energyrelated CO₂ emissions utilizing optimized support vector machine. Journal of Cleaner Production 250: 1–13.

22.

, et al. (2018) CVaR risk-based optimization framework for renewable energy management in distribution systems with DGs and EVs. Energy 143: 323–336.

23.

Hou

Zhu

, et al. (2021) Mid-term prediction of electrical energy consumption for crude oil pipelines using a hybrid algorithm of support vector machine and genetic algorithm. Energy 222: 1–13.

24.

Yan

(2018) Optimal portfolio of continuous-time mean-variance model with futures and options. Optimal Control Applications and Methods 39: 1220–1240.

25.

Yang

(2017) Portfolio optimization of Chinese overseas oil and gas assets based on CVaR. Master Dissertation, China University of Petroleum, Beijing, China.

26.

Chen

, et al.(2022) Optimal portfolio of a 100% renewable energy generation base supported by concentrating solar power. Renewable and Sustainable Energy Reviews 170: 1–15.

27.

Zhang

(2007) Multi-business combination decision model. Master Dissertation, Northwest University of Technology, Xi’an, China.

28.

Zhang

Tang

Chen

(2021) Multi-step-ahead crude oil price forecasting based on two-layer decomposition technique and extreme learning machine optimized by the particle swarm optimization algorithm. Energy 229: 1–13.