Proceedings of the Institution of Mechanical Engineers,Part I: Journal of Systems and Control Engineering

A scheme for an electrohydraulic vibrator excited by a two-dimensional valve is proposed, which significantly extends the frequency range compared to that of vibrators excited by conventional servo valves. In the two-dimensional valve, the rotary motion of the spool coordinates the relative motion of the grooves on the spools area with respect to the windows on the sleeve which in turn, alternates the oil flowrate into and out of the chambers of the hydraulic cylinder (motor) and subsequently excites the piston (rotor) to vibrate. The linear motion of the two-dimensional valve spool is used to vary the peak flowrate and thus the amplitude of the output vibration. The frequency of the vibration excited by the two-dimensional valve is related to the rotary speed of the spool and the number of the grooves on the spool area (windows on the sleeve). This configuration extends the frequency of the hydraulic vibrator by increasing the number of the grooves on the spool area (windows on the sleeve) and the rotary speed of the spool which are in the perfect lubrication of oil.

A model of the two-dimensional vibrator is developed and compared to its experimental counterpart. The vibration excited by the two-dimensional valve is influenced by pressure saturation, the elastic force, and the hydraulic resonance. There is a critical valve linear opening, beyond which the output force (torque) reaches a saturation value in both a positive and negative sense. Both simulation and experimental results show that at lower frequencies the ascent and descent slopes of the output force show some inconsistency which becomes more significant above the critical valve linear opening but drops off with a reduction in the valve linear opening. In a higher frequency range, the vibration excited by the two-dimensional valve is mainly influenced by hydraulic resonance. As the input frequency approaches the hydraulic resonant frequency, the output excited vibration essentially becomes the hydraulic resonance. Therefore, the effective frequency range of the hydraulic vibration is not only decided by the frequency bandwidth of the two-dimensional valve, but is influenced by the hydraulic resonance. Nevertheless, the study does provide an access to the high-frequency excitation of the hydraulic vibration.

A tailored non-linear fluctuation smoothing rule is proposed in this study to improve the performance of scheduling jobs in a semiconductor manufacturing factory. The tailored non-linear fluctuation smoothing rule is modified from the well-known fluctuation smoothing rules with three innovative treatments. At first, the fluctuation in the remaining cycle time estimate is smoothed and then its influence is balanced with that of the release time or the mean release rate. Subsequently, the difference in the slack is magnified by applying the ‘division’ operator instead. Thirdly, the content of the non-linear fluctuation smoothing rule can be tailored to the semiconductor manufacturing factory to be scheduled. In order to evaluate the effectiveness of the proposed methodology, production simulation is also applied in this study to generate some test data. According to experimental results, the proposed methodology out-performed nine existing approaches in reducing the cycle time averages and standard deviations. In addition, the tailored non-linear scheduling rule was also shown to be a Pareto optimal solution for scheduling jobs in a semiconductor manufacturing factory.

The success of model-based control of chemical processes is dependent on good process models. For processes that are poorly known, the generic modelling capability of neural networks offers an attractive alternative. However, for satisfactory performance, the conventional implementations of neural networks require large sets of offline data in addition to online measurement of key variables, such as concentrations. Meeting each of these requirements is often infeasible in chemical processes. By combining the structural information from a first-principles model and the virtual supervisor-artificial immune algorithm, a novel hybrid neural network, called a structure approaching hybrid neural networks (SAHNN), is proposed. The proposed approach solves the structural problem of neural models and requires a more manageable number of offline data and online key variables. The accurate prediction of online partially unmeasurable concentrations in a batch reactor demonstrated that SAHNN is a promising tool to model complicated batch processes and can be utilized as a vehicle for the control and optimization of other similar chemical reactors.

A qualitative approach is introduced to reason the spatial configuration of mechanisms in the conceptual design phase. Using the qualitative vector of fuzzy information, position vector, and qualitative constraint matrix, a reasoning structure for the spatial configuration of the mechanism is presented based on qualitative sign algebra. Assuming that complex mechanisms are composed of combinations of basic components, the reasoning structure is applied to reason about the spatial configuration of a complex mechanism. An analysis of the kinetics of the system is performed by reasoning with the vectors of basic components. An analysis of dynamic performance is presented.

In this paper, a technique for computing robust controller for multivariable time-invariant linear systems via state-derivative feedback is introduced such that the sensitivity of the closed-loop system eigenvalues to perturbations in the system and gain matrices is minimized. The proposed technique can be applied for any controllable system with some restrictions when assigning zero poles. This article focuses on the derivation of feedback gain when the open-loop state matrix is singular. The available degrees of freedom offered by state-derivative feedback for multivariable system are utilized to improve robustness of the closed-loop system. Finally, a computational algorithm and a numerical example are introduced to demonstrate the effectiveness of the proposed technique.

The stability issue of mobile manipulators, particularly when the end-effector and the vehicle have to follow a predefined trajectory (for some special duties like painting a plane or carrying a light load), is a crucial subject and needs special attention. In this paper, by utilizing the manipulator compensation motions, the instantaneous proper configuration for a redundant mobile robot is determined. A fast methodology taking into account the dynamic interaction between the manipulator and the vehicle is proposed for enhancing the tipover stability (i.e. stability against overturning) of the mobile manipulator by employing the soft computing approach including a genetic algorithm, neural network, and adaptive neuro-fuzzy inference (ANFIS) controller. A genetic algorithm is utilized to find a minimum value for a function, called the performance index here, to determine the tipover stability measure. In such an algorithm, utilizing the values attained, a neural network look-up table is planned. This look-up table will expedite the operation of tracking a specified trajectory for the end-effector while the mobile manipulator simultaneously maintains its tipover stability in the most favourable manner. The benefit of using the neural network is its capacity for online control; the genetic algorithm alone cannot provide this. In this case the performance index is minimized and the rule base (the main part of an ANFIS controller) for the adaptive neuro-fuzzy controller is designed. It is considered that the tipover stability of the mobile manipulator is increased by applying the rule base of the ANFIS controller to the system. For evaluating the effectiveness and performance of the proposed algorithm, a spatial mobile manipulator is examined.

This study presents a novel feedback linearization control of non-linear multi-input multi-output uncertain systems for the tracking and almost disturbance decoupling performances. The main contribution of this study is to construct a controller, under appropriate conditions, such that the resulting closed-loop system is valid for any initial condition and bounded tracking signal with the following characteristics: input-to-state stability with respect to disturbance inputs and almost disturbance decoupling. In addition, a new theorem on robust stability is proposed in this study to provide a new criterion for closed-loop stability. A typical case, which cannot be solved by any previous study on the almost disturbance decoupling problem, is proposed in this study to exploit the fact that the tracking and the almost disturbance decoupling performances can be easily achieved by the proposed approach. Finally, the proposed control law is simulated in a half-car active suspension system on which the effectiveness of the design is verified.

The optimal energy efficient operating conditions of a solid oxide fuel cell (SOFC) system needs to be studied by considering the whole closed circuit system rather than a standalone study of the cell efficiency. This study is performed in this paper with the aid of steady state models of the SOFC, the after-burner, and the heat exchanger. For the first time, a comprehensive steady state model of the SOFC is developed and validated. A recursive algorithm with two cascaded optimization loops is applied to maximize the SOFC system efficiency and also to obtain the corresponding cell operating conditions. The developed steady state model aids in the implementation of the optimization procedure. Controlling the SOFC system for maximum efficiency operation for variable loads requires complex control laws. However, it is found that an appropriately chosen constant fuel utilization (FU) operation closely approximates the maximum efficiency operation of the fuel cell in its operating range. This is validated through closed-loop dynamic simulations of a bond graph model of the complete SOFC system. Three different commonly used control strategies and their implications on the energy and exergy efficiencies of the system, as well as their transient load following capabilities, are investigated.

This paper deals with the partial pole placement design of a proportional-integral (PI) observer. The proposed partial pole placement approach offers a new method for designers to assign the PI observer gains on a detectable system. Analysis and design of the PI observer are discussed in detail. A crucial lemma and discussions to the proposed approach are presented, and an illustrative numerical example is given to verify the proposed design approach.

The method introduced here is applicable for multi-input multi-output, linear, and time-invariant systems. The state and output equations of the system, which are originally expressed in the t-domain, are first transformed into the s-domain. Then, input-output decoupling is achieved by generating the actual control variables as combinations of virtual control variables in such a way that each output is controlled by only the dedicated one of the virtual control variables. As the next stage, appropriate linear controllers are designed to generate the virtual control variables with feedback, feedforward, and disturbance rejection features. The design method is algebraic and involves the assignment of not only the closed-loop poles but also the type numbers associated with the reference and disturbance inputs. The multivariable transformation filter that converts the virtual control variables into actual ones is formed properly to prevent any possibility of internal instability.

This paper describes a rule-based neuro-fuzzy technique for the navigation of 1000 robots in a cluttered environment. The planning and coordination between the mobile robots is extremely difficult. In the present analysis rule-based and rule-based neuro-fuzzy techniques are used to navigate multiple mobile robots in unknown and partially known environments. The aim of the robots is to reach a predefined goal. Based upon a reference motion, direction, distances between the robots and obstacles, and distances between the robots and targets, different types of rules are defined heuristically and refined later to find the steering angle. The control system combines a repelling influence related to the distance between robots and nearby obstacles and an attracting influence between the robots and targets. Then a hybrid rule-based neuro-fuzzy technique is analysed to find the steering angle of the robots. Simulation and experimental results show that the proposed rule-based neuro-fuzzy technique shows a better navigation performance in complex and unknown environments than the simple rule-based technique.