ANN Optimization of Weighting Factors Using Genetic Algorithm for Model Predictive Control of PMSM Drives

Yao, Chunxing; Sun, Zhenyao; Xu, Shuai; Zhang, Han; Ren, Guanzhou; Ma, Guangtong

doi:10.1109/tia.2022.3190812

Cited by 35 publications

(10 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Weighting factor removal by reference transformation 33,34 Higher computational burden as compared with conventional PTC and difficult to incorporate multiple control objectives 35 Weighting factor tuning based on coefficient of variation 36 Optimized weights are uncertain in this method and complex calculations are required to implement on hardware Weighting factor tuning based on TOPSIS and NSGA-II methods 37 TOPSIS and NSGA-II algorithms require complex calculations leading to computational challenges 12 Weighting factor removal by Ranking method 38 Ranking based techniques become unfeasible as number of control objectives increases 39 Tuning of weighting factor based on simple additive technique 40 Although technique is simple but not suitable for multiple control objectives 11 Weighting factor tuning based on current ripples 41 Highly dependent on parameter estimation 8,42 Tuning of weighting factor based on error of control objectives 43 This method becomes challenging and complex when number of control objectives increases 44 Weighting factor tuning using Genetic Algorithm (GA) 45 , Simulated Annealing (SA) 42 or Gravitational Search Algorithm (GSA) 43 , Artificial Neural Network 46 , Ant colony based optimization 47 These algorithms are very complex and pose computational challenges 48 Weighting factor tuning based on algebraic/numerical techniques 49 Design complexity increases as slection of weighting factor increases 50 Weighting factor selection based on homogeneous cost functions [51][52][53] This technique is relatively efficient but unable to include multiple control objectives 54 Direct vector selection based techniques to remove weighting factors from cost function 55,56 Direct vector selection techniques provid lower computational burden and lower complexity , however cannot incorporate multiple control objecitve 57 Weighting factor elimination by using cascaded structure of FCS-MPC …”

Section: Ptc Methods Limitationsmentioning

confidence: 99%

See 1 more Smart Citation

Optimal weighting factor design based on entropy technique in finite control set model predictive torque control for electric drive applications

Shahid,

Jin,

Abbasi

et al. 2024

Sci Rep

View full text Add to dashboard Cite

In the conventional finite control set model predictive torque control, the cost function consists of different control objectives with varying units of measurements. Due to presence of diverse variables in cost function, weighting factors are used to set the relative importance of these objectives. However, selection of these weighting factors in predictive control of electric drives and power converters still remains an open research challenge. Improper selection of weighting factors can lead to deterioration of the controller performance. This work proposes a novel weighting factor tuning method based on the Multi-Criteria-Decision-Making (MCDM) technique called the Entropy method. This technique has several advantages for multi-objective problem optimization. It provides a quantitive approach and incorporates uncertainties and adaptability to assess the relative importance of different criteria or objectives. This technique performs the online tuning of the weighting factor by forming a data set of the control objectives, i.e., electromagnetic torque and stator flux magnitude. After obtaining the error set of control variables, the objective matrix is normalized, and the entropy technique is applied to design the corresponding weights. An experimental setup based on the dSpace dS1104 controller is used to validate the effectiveness of the proposed method for a two-level, three-phase voltage source inverter (2L-3P) fed induction motor drive. The dynamic response of the proposed technique is compared with the previously proposed MCDM-based weighting factor tuning technique and conventional MPTC. The results reveal that the proposed method provides an improved dynamic response of the drive under changing operating conditions with a reduction of 28% in computational burden and 38% in total harmonic distortion, respectively.

show abstract

Section: Ptc Methods Limitationsmentioning

confidence: 99%

“…Other meta-heuristic methods include gravitational search algorithm (GSA) 43 , non-dominated sorting genetic algorithm II (NSGA-II) and Particle Swarm Optimization (PSO) [10][11][12] . ANN methods are reported in 46,[63][64][65][66] . However, they also suffer from higher computational burden and tend to increase the complexity of the controller.…”

mentioning

confidence: 99%

Optimal weighting factor design based on entropy technique in finite control set model predictive torque control for electric drive applications

Shahid,

Jin,

Abbasi

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…A summary of the most popular techniques is presented in Figure 3. It is commonly known that the number of training epochs defined for the calculations significantly influences the fitting of data using the neural model [23]. Increasing the number of iterations results in an improved agreement with the training data.…”

Section: Preliminaries and Short Description Of Methodologymentioning

confidence: 99%

Neural Network Applications in Electrical Drives—Trends in Control, Estimation, Diagnostics, and Construction

Kaminski

Tarczewski

2023

Energies

View full text Add to dashboard Cite

Currently, applications of the algorithms based on artificial intelligence (AI) principles can be observed in various fields. This can be also noticed in the wide area of electrical drives. Consideration has been limited to neural networks; however, the tasks for the models can be defined as follows: control, state variable estimation, and diagnostics. In the subsequent sections of this paper, electrical machines, as well as power electronic devices, are assumed as the main objects. This paper describes the basics, issues, and possibilities related to the used tools and explains the growing popularity of neural network applications in automatic systems with electrical drives. The paper begins with the overall considerations; following that, the content proceeds with the details, and two specific examples are shown. The first example deals with a neural network-based speed controller tested in a structure with a synchronous reluctance motor. Then, the implementation of recurrent neural networks as state variable estimators is analyzed. The achieved results present a precise estimation of the load speed and the shaft torque signals from a two-mass system. All descriptions in the article are considered in the context of the trends and perspectives in modern algorithm applications for electrical drives.

show abstract

“…Some other efforts devoted to removing weighting factors need to normalize controlled quantities or convert them into equivalents [8]. It is interesting that some advanced algorithms such as artificial neural networks [9], fuzzy optimization [10], genetic algorithms [11], and particle swarm optimization [12], have been used to deal with the weighting factor issue in MPC. These advanced algorithms increase the complexity, which removes one of the advantages of MPC implementations.…”

Section: Introductionmentioning

confidence: 99%

Alternating Sequential Model Predictive Control of Matrix Converter

Zhang

Rivera

Wheeler

2023

2023 IEEE International Conference on Predictive Control of Electrical Drives and Power Electronics (PRECEDE)

View full text Add to dashboard Cite

Weighting factor design is one of the challenges faced by model predictive control (MPC) when multiple control objectives are to be achieved simultaneously. This paper proposes an alternating sequential MPC (ASMPC) in which the cost functions are evaluated in alternating order, avoiding the weighting factor selection and the need to determine the evaluation order for each cost function. Reduced computational burden is achieved at the same time. The proposed ASMPC has been applied to the control of a matrix converter and simulation results verify the feasibility and effectiveness of the proposed controller. The proposed ASMPC can be readily adapted and used for other power electronic converters.

show abstract

ANN Optimization of Weighting Factors Using Genetic Algorithm for Model Predictive Control of PMSM Drives

Cited by 35 publications

References 39 publications

Optimal weighting factor design based on entropy technique in finite control set model predictive torque control for electric drive applications

Optimal weighting factor design based on entropy technique in finite control set model predictive torque control for electric drive applications

Neural Network Applications in Electrical Drives—Trends in Control, Estimation, Diagnostics, and Construction

Alternating Sequential Model Predictive Control of Matrix Converter

Contact Info

Product

Resources

About