A BPSO based collaborative management platform for multi-source data security in power grids

3.1 Basic application platform based on XML configuration

The application configuration of multi-source data security collaborative management of digital power grid is realized by using XML Extensible Markup Language, and the tagged electronic file of multi-source data security collaborative management of digital power grid is constructed by using a subset of standard generic markup language, so that it has a structural markup language. In the information symbol model of multi-source data security collaborative management in digital power grids, through this labeling, the feature quantity of multi-source data security collaborative data in digital power grids is calculated. It can be used to label data and define data types, and is a source language that allows users to define their own markup language. It is very suitable for the security and collaborative management of multi-source data of digital power grid in the World Wide Web transmission. It provides a unified method to describe the exchange information of the security and collaborative management of multi-source data of digital power grid. Through dynamic information interaction, it makes the security and collaborative management of multi-source data of digital power grid independent of structured data of applications or suppliers. It uses the Internet’s document information transmission control technology to improve the stability of system output [11].

3.2 B/S architecture

Develop system software hierarchy under the B/S architecture, which is a cross platform application software structure that supports all software and hardware systems of TCP/IP protocol. The hybrid development of the digital power grid multi-source data security collaborative management platform is carried out using the Native app local application program and the Web app web application program. The software development of the platform mainly includes the development of the platform’s power enterprise information database, the development of the data processing module of the digital power grid multi-source data security collaborative management platform, the development of the network communication module, and the development of the digital power grid multi-source data management output interface. The network design of the digital power grid multi-source data security collaborative management platform is carried out under the B/S structure system, and the logical database design is carried out with MySQL as the database, which reduces the time and energy for maintaining the client software and customer training, and also facilitates the use of users.

3.3 Component based design

Using the logical component structure design method, a component control model for the digital power grid multi-source data security collaborative management platform is established. A distributed computing framework design scheme based on Java combines reusable object-oriented components with a healthy and reliable implementation environment to submit multi application collaboration. By combining mobile communication networking and IoT technology, the network networking control of the digital power grid multi-source data security collaborative management platform is carried out. Based on the dynamic management method of application programs, these business components can be independently installed and upgraded. The advantage of its relationship lies in its ability to tightly integrate or insert selected third-party product components. Users can obtain a set of components to meet their specific and unique maintenance, scheduling, and procurement needs [12].

3.4 J2EE technology and spring framework

Adopting the Browser/Server architecture method, combined with J2EE technology, the server-side design of the Java technology development for the digital power grid multi-source data security collaborative management platform is implemented. The J2EE application server is used to establish a J2EE enterprise level deployment platform for the digital power grid multi-source data security collaborative management platform. Due to their adherence to the J2EE specification, enterprise level applications developed using J2EE technology can be deployed on various J2EE application servers.

In the design of a digital power grid multi-source data security collaborative management platform, Java EE 5 design method is used to implement some important features to simplify the work of developers. The JCP organization has summarized the experience and lessons learned from past J2EE practices and made significant adjustments to J2EE technology in the Java EE 5 specification. In Java EE, support for web services remains an important topic. Web services in the Java Enterprise Edition platform mainly revolve around two application development interfaces: the Java API (JAX-WS) 2.0 for XML Web services, which is a continuation of JAX-RPC and JAXB.

In the collaborative management of multi-source data security in digital power grids, Spring’s open source framework design method is adopted, which is created to address the complexity of enterprise application development. Spring uses basic JavaBeans to achieve an EJB-based hierarchical architecture for collaborative management of multi-source data security in digital power grids. It uses 7 core container networking control technology to create, configure, and manage the collaborative management of multi-source data security in digital power grids. The physical architecture of the network structure layer for collaborative management of multi-source data security in digital power grids is shown in Fig. 2.

OPEN IN VIEWER

Fig. 2

Construction of Network Physical Architecture.

This article adopts middleware technology to achieve heterogeneous hierarchical data fusion in the digital power grid multi-source data security collaborative management network. The middleware designed in this article for the digital power grid multi-source data security collaborative management network includes: (1) network middleware of Hibernate framework; (2) Configure middleware; (3) Hibernate is an open source object relationship mapping framework that enables Java program scheduling and object compilation methods to manipulate digital power grid multi-source data security collaborative management network databases. Hibernate has a total of 5 core interfaces, namely Session, SessionFactory, Transaction, Query, and Configuration. These 5 core interfaces will be used in any development in the digital power grid multi-source data security collaborative management platform. Through these interfaces, not only can persistence scheduling and object management of multi-source data resources of the digital power grid be realized, but also transaction control can be carried out through access to the keywords “provides” and “uses", and the Hibernate framework structure of the digital power grid multi-source data security collaborative management platform can be obtained, as shown in Fig. 3.

OPEN IN VIEWER

Fig. 3

Hibernate framework structure of collaborative management platform.

4.1 Construction of scheduling model for discrete particle swarm optimization algorithm

Firstly, the discrete particle swarm optimization algorithm is a heuristic optimization algorithm that seeks the optimal solution in the solution space by simulating the behavior of bird populations. For the scheduling problem of multi-source data fusion, the scheduling tasks of different data sources can be viewed as individual solutions in the solution space. The discrete particle swarm optimization algorithm iteratively updates the position and velocity of particles to gradually optimize and solve the objective function.Build a secure cooperative scheduling of multi-source data of digital power grid based on discrete particle swarm optimization algorithm. There are m particles to form a distributed population of multi-source data of digital power grid. The selected individuals need to be fused with multi-source data of digital power grid through crossover and mutation operations to generate individuals with higher fitness, The digital grid multi-source data security cooperative control particle updates itself by updating two “extreme values": one is the optimal solution found by the particle itself, which is called the individual extreme value pbest of digital grid multi-source data security fusion, and designs a population’s fitness function, genetic operations, including crossover and mutation, convergence conditions The particle updates its speed and position according to the following two formulas:

v t = wv t - 1 + c 1 rand 1 ( ) · ( pbest - x t - 1 ) + c 2 rand 2 ( ) · ( gbest - x t - 1 )(1)

x t = x t - 1 + v t(2)

Wherein, v _t is the current speed of particles in the process of multi-source data fusion of digital power grid, and x _t is the fitness value of persistent scheduling of multi-source data resources of power grid. c₁ and c₂ are operator constants, usually taken as c₁ = c₂ = 2; rand₁ () and rand₂ () are random numbers for program scheduling and object compilation between [0,1]. For each function corresponding to Xi, they are:

l i ( k ) = ( 1 - ρ ) l i ( k - 1 ) + γ f ( x i ( k ) )(3)

Wherein, f _i is Xi fitness function, and P ij ( k ) represents the probability of the i-th particle moving to the j-th particle scheduled by its grid multi-source data security cooperative management network heterogeneous neighbor set at the moment. The least square problem is solved to optimize the security cooperative management control group meme group of power grid multi-source data, sort all particles according to fitness, and obtain the information entropy of power grid multi-source data security cooperative management control through Lagrange deformation processing:

min Q 1 2 ∥ Q | Ω - P | Ω ∥ F 2 + μ ∥ Q ∥ *(4)

The particle mutation extremum model for secure collaborative scheduling of power grid multi-source data is obtained through the meme group jump cycle method, and the operator value is obtained. Under the unconstrained condition of non independent related variables, the process entropy of secure collaborative management of power grid multi-source data is obtained, which satisfies:

∥ Q | Ω - P | Ω ∥ F 2 ≈ # ( Ω ) σ ˆ 2(5)

Set μ = ( n 1 + n 2 ) p σ ˆ , n₁ and n₂ are block matrices used to represent the distance between particles. Based on the update iteration order in the meme group, the distance between individuals in the power grid multi-source data security collaborative control particle group is obtained as:

dist ( i , j ) = ∑ k = 1 d ( x ik - x jk ) 2 + ( f ( X i ) - f ( X j ) ) 2(6)

Considering the global optimization problem min f (x), the discrete particle swarm optimization algorithm is used to search for the optimal value in the D dimension space using a group of particle sets, and the variation fitness value of the grid multi-source data security collaborative management is obtained as follows:

MdistFg = ∑ i = 1 N dist ( i , F g ) / N(7)

Wherein, F _g is the association rule function, which has not yet been updated during the search process to achieve particle adaptive optimization for secure collaborative scheduling of multi-source data in the power grid.

The hybrid ICMICPSO-BP algorithm is used to conduct secure cooperative scheduling of power grid multi-source data, and FNNs are trained. Assuming that the optimal movement probability of secure cooperative scheduling of power grid multi-source data is P ij best ( k ) , and the movement step size is S, the position of the particle of the improved particle swarm optimization at the k + 1 time of the grid multi-source data secure cooperative scheduling subspace is:

x i ( k + 1 ) = x i ( k ) + s ( x j ( k ) - x i ( k ) ∥ x j ( k ) - x i ( k ) ∥ )(8)

The gradient descent (BP) method based on error back propagation is used to train the grid multi-source data security cooperative scheduling process, and the mean square error function of particles is calculated as:

E = ∑ j = 1 q E j / E j ( q * k ) w h e r e E j = ∑ k ε k 2 = ∑ k ( d k − c k ) 2(9)

In the formula, q is the number of input samples for multi-source data in the power grid, ɛ _k is the initial weight sensitivity coefficient of the output layer, and d _k is the global extreme value for the security collaborative control of multi-source data in the k-th node of the output layer. The differentiated spectral analysis method is used to test the PSO algorithm to search in the D dimensional space, and m grid multi-source data security cooperative scheduling particles form a population. Then the current position of the ith particle can be expressed as the vector xi = (xi1, xi2,..., xiD), and its speed can be recorded as the vector vi = (vi1, vi2,..., viD). Considering the global optimization problem, the grid multi-source data security cooperative scheduling training can be carried out through error back-propagation gradient decline, The particle update formula is obtained as follows:

{ v i d t + 1 = ω v i d t + c 1 r 1 ( p i d − x i d t ) + c 2 r 2 ( p g d − x i d t ) x i d t + 1 = x i d t + v i d t + 1(10)

The discrete particle swarm optimization update formula for collaborative management and fusion of multi-source data security in the power grid is:

{ x id t + 1 = ω x id t + c 1 r 1 ( p ad - x id t ) + c 2 r 2 ( p gd - x id t ) p ad = 1 m ∑ i = 1 m p id(11)

Wherein, pad is the average value of the individual optimal position of the grid multi-source data security collaborative control, the probability density function of the optimal clustering solution that tends to be stable, and the particle set of the grid multi-source data security collaborative filtering. Taking this as the input weight value, the particle swarm optimization with dynamic inertia weight is used to obtain the global optimal Xex best of grid multi-source data security cooperative scheduling. Particles update the speed and position of grid multi-source data security cooperative control according to individual optimal and global optimal, C1 and C2 are learning factors, rand() is a random function based on [0,1] uniform distribution, The particle algorithm update formula for secure collaborative scheduling of multi-source data in the power grid is as follows:

e a = ( W β − W α ) / s u m ( 1 : I t e r a t i o n s )(12)

ω = ω - ( Iterations - iter ) × ea(13)

Wherein, W α, w β The upper and lower limits of inertia weight represent the secure collaborative scheduling of multi-source data in the power grid. Based on the above algorithm design, discrete particle swarm optimization algorithm is adopted to achieve multi-source data fusion and security collaborative management of the digital power grid.