Optimization is often encountered in many fields such as engineering technology, theoretical research and economic management. In recent years, with the rapid development of modern control theory and computer technology, the application of optimization theory and technology has become increasingly widespread, and has achieved great economic and social benefits. The motor optimization design that began in the early 1960s was one of them. One of the hot spots.

The so-called motor optimization design refers to a technology that satisfies national standards, user requirements, and specific constraints, so that the motor performance, volume, power, weight and other design performance indicators can be optimized. It can be described as a complex and constrained. Some progress has been made in the optimization design of nonlinear, mixed-discrete multi-objective programming problems.

The earliest foreign countries applied the classical extreme value theory to the motor optimization design program. The domestic Tsinghua University used the hybrid discrete planning method to design a special high-efficiency special three-phase asynchronous motor earlier. The main indicators reached the advanced level of similar products of the world at that time. Since then, many domestic scholars have conducted a lot of exploration and research on nonlinear optimization algorithms that can be used to optimize the motor design. There are many successful examples. Using Powell's method and supplemented by a global optimization technique-filled function method, the problem of optimal design of single-phase motor core series is solved. Through comparative research, it is believed that the random search method has the advantages of simplicity, validity, applicability, and uses this algorithm to develop a CAD software package for single and three-phase, synchronous and asynchronous motors and exciter optimization design. In this paper, the cost of the motor is taken as the objective, and the new method of hybrid discrete optimization based on the generalized coordinate rotation method is used to solve the stator chip optimization problem. The comparative study shows that the generalized coordinate rotation method is a better method to solve the hybrid discrete optimization problem. Through the direct combination of the direct method and the indirect method, the dimension reduction method and the sequential decomposition method are adopted for the optimization of the small-sized cage motors. The experimental data show that the sequential decomposition method is superior to the dimension reduction method. The problem of discrete variables and total extremum problems are studied in detail. Taking the optimization design of large hydroelectric generators as an example, a total extremum algorithm suitable for the optimal design of the motor is verified. In addition, many other algorithms, such as simplex method and complex method, have been applied in the optimization design of the motor. Due to space limitations, they are not listed one by one.

How to determine the motor optimization model and improve the optimization algorithm to meet the motor's specific application performance requirements is a common concern for motor engineers. Although the traditional optimization design strategy has achieved certain results in the practice of guiding motor design, there are still many problems, such as the optimization result is closely related to the choice of the initial solution; the optimization algorithm often converges to the local extreme point near the initial value, it is difficult The ideal global optimal results are obtained. In addition, the discrete variable optimal design is a multidimensional optimization problem in the multi-mode space with noise. The traditional analytical and numerical methods no longer exist in the fine features of the objective function. Therefore, the traditional optimization method cannot easily or accurately complete the motor multidimensional multimode optimization task.

3 Introduction to modern motor optimization design Modern heuristic optimization algorithms such as tabu search (TS) and simulation evolution (SE) have been rapidly developed. In order to further improve the motor design level, motor workers began to study these new optimization theories and motor design techniques. Combination and crossover, and gradually formed a modern motor optimization design technology that can achieve the overall optimal. Here are some of the several typical and representative optimization design techniques for analysis.

3.1 Genetic Algorithm Based Motor Optimization Design Genetic Algorithm (GA) is a bionic algorithm that simulates life evolution: its operation object is a group of feasible solutions, rather than a single feasible solution; there are multiple search tracks instead of a single one. It has good parallelism; it only needs to use the target's value information and does not need high-value information such as gradients, so it is suitable for optimization of any large-scale, highly non-continuous discontinuous multi-peak functions and optimization of objective functions without analytical expressions. , has a strong universality; its selection mechanism is a "soft" choice, coupled with good parallelism, so that it has a good global optimization and robustness; the feasible solution set for operations is encoded The objective function is interpreted as the adaptive value of the coded individual and thus has good operability and simplicity.

The amount of processing is also difficult to determine... Actually, the modelling of the motor is particularly useful in combination with tactile shifts. Consider the optimization point. again. After netbookmark3, GA has become the focus of professional research and is widely used in the optimization of various motors. Has been widely used in household appliances, power tools, medical equipment and light industrial equipment in the single-phase motor has a wide range of large-scale, variety, fast replacement and other characteristics, its optimization design is undoubtedly has great commercial value and broad market prospects . Wang Jie et al. optimized the genetic algorithm and generalized coordinates using the generalized coordinate rotation method. The discrete variables are optimized by the extension and contraction method along the axis, while the continuous variables are optimized by the Powell method with good quadratic convergence. From the optimization of five single-phase motors, the cost is significantly reduced, and the number of iterations of the algorithm is reduced, and the optimization efficiency is improved. In order to improve the overall quality of the initial solution group, Wu Xinzheng accelerated the optimization process, added the original scheme before the single-phase motor optimization to the initial solution group, improved the crossover operator, and improved the convergence of the ordinary GA algorithm. .

The research on the optimal design of a single motor has achieved good results. However, the optimized design of the series motor is more practical than the stand-alone optimization. Although there has been some progress in this aspect of the work, due to the limitations of the optimization theory, it is basically still in the preliminary experimental stage. For the optimal design of a series of motors, each sub-goal is generally an effective cost of the motor, and a large output motor plays a decisive role in the series. Wu Derong et al. performed linear weighting on the objective function, transformed the multi-objective optimization into a single-objective programming problem, and directly used the method of solving single-objective planning to complete the optimization design of the Y-series motor. At the same time, taking the genetic algebra and the solution quality as the convergent termination condition of the genetic iteration avoids the problem of premature convergence of GA to the local optimal solution due to the use of a single convergence criterion.

In recent years, permanent magnet motors have been widely used in high-performance drive systems because of their superior electromagnetic properties. However, due to the complex geometry of permanent magnet motors, and the special performance and high cost of permanent magnet materials, the research on optimization design of permanent magnet motors is particularly important.

The 5kW rare earth permanent magnet shielded motor is optimized. After the product is applied to the shielded electric pump, the head and flow rate are greatly improved. Because of the complexity of the calculation of the performance of the switched reluctance motor (SR) and the strong coupling between the control parameters, the application of the optimization algorithm in such motors is difficult. Wu Jianhua used GA to study the optimization design of SR motor more comprehensively, and discussed how to optimize the characteristics of the SR motor in the optimization process. Brushless motor (BLDCM) as a self-controlled commutation permanent magnet motor has good start and speed regulation characteristics. The motor body has a simple structure, high power density, and a wide application prospect. At present, due to the high price of NdFeB permanent magnet materials and the high motor cost, it is hoped that the cost will be reduced through optimized design and the cost performance of the motor will be improved. Shi Shan et al. proposed an improved adaptive genetic operator-genetic operator that automatically changed with fitness value, adopted a larger genetic operator value for individuals far from the optimal value, and adopted a smaller genetic value for individuals close to the optimal value. Operator value. Actual results should be based on the genetic algorithm to meet the performance of the premise under the conditions to obtain a good optimization effect, the probability of obtaining the global optimal solution has improved significantly before the improvement. The linear motor is a novel motor that directly converts electrical energy into linear motion without any intermediate switching device. It is widely used in the fields of production, processing and transportation due to its linear running characteristics. Linear motors have evolved from rotating electrical machines. Although the two are completely identical in principle, due to different structures and uses, there are great differences in the optimal design. Selecting the electromagnetic thrust and volume of the permanent magnet linear synchronous motor as the optimization goal, using the penalty function method to convert the constraint problem into no constraint, to overcome the problem of early convergence (premature) of the GA algorithm, increase the mutation probability appropriately, and eliminate part of the old individuals. A viable new individual, and an experiment with a 20-slot single-sided, short primary length secondary permanent-magnet linear synchronous motor (NdFeB permanent magnet NTP264H) experiment proved the effectiveness of the algorithm. Qiu Lin and others established a three-dimensional finite element numerical analysis and GA combined linear synchronous motor optimization design model to optimize the linear propulsion of the motor to improve the smoothness of the linear motor operation in the traction system. The paper also optimizes the harmonic components under the premise of unchanged motor output. The results show that the combination of GA and finite element is an effective global optimization method in the field of motor optimization design.

In summary, GA and its improved algorithms can solve almost all problems in the field of motor optimization design, and have produced huge economic benefits. However, in practical applications, we find that GA optimization techniques are not perfect and there are still many areas for improvement: initial populations are randomly generated, and when the solution groups are not evenly distributed, they tend to fall into a local optimum; when the population evolves to a certain algebra Individual concentration is too high to maintain the diversity of individuals and cause immature convergence. Because motor optimization is a hybrid discrete optimization problem with both continuous and discrete variables, GA is characterized by the need to artificially discretize continuous variables. . If the best point is not at an artificial discrete point, the optimization point can never be found; in order to make the optimization run in the entire solution space, the matching set N must be made large enough, which increases the calculation time, if the matching set gets too small, It is possible to lose the opportunity to find the best of the world.

2The optimization design of the motor based on immune algorithm shows that in the optimization design of the motor, the modified self-adaptive publ mechanism of the real organism, if the neural network is so far, there are many non-deterministic optimizations. The method (ie, stochastic optimization method) is applied to motor optimization design.Compared with deterministic algorithm, the advantage of non-deterministic algorithm is that it has more chances to obtain the global optimal solution.Many non-deterministic algorithms mostly embody human beings. Simulation of the brain's information processing mechanism, the fuzzy system is the simulation of human's way of thinking, and the genetic algorithm is the simulation of the biological survival evolution. Then the recently developed immune algorithm (IA) has become an emerging simulation of the biological immune mechanism. For the optimization method, the affinity has two meanings: on the one hand, the relationship between the antibody and the antigen, ie the degree of matching between the solution and the target; on the other hand, the relationship between the antibodies is explained, ie the control is applicable to the antigen (target). The excessive production of the same antibodies guarantees the diversity of candidate solutions. The role of calculating the affinity is to use a set of memory cells for preservation A set of antibodies to the antigen (candidate solution to the optimization problem), so that for the antigens that have ever appeared, IA produced the corresponding antibodies more quickly than before. It can be seen that both GA and IA are bionic optimization models, but both The optimization mechanism is not the same, and IA has advantages that other algorithms cannot match: IA has a program for calculating affinity that can reflect the self-regulating function of the real immune system and ensure the diversity of solutions; IA conducts memory training to ensure that the optimization process can be rapid Convergence to the global optimal solution.

At present, in the optimization calculation, IA and GA are combined more and more to complement each other, forming a compound optimization calculation model based on immune genetic mechanism. It can be seen that the immune genetic algorithm has been widely applied to all aspects of modern engineering design problems. Although applications in the field of motor design are rare, it can be determined that this optimization algorithm also has high application value for motor optimization design.

3 Other new motor optimization design algorithm overview In the process of motor optimization design, the feasibility of the initial value of the design variables is the key to the optimization design success or failure. Based on the expertise and experience provided by the motor design experts, the combination of expert system and traditional optimization design provides an advantageous way to solve the feasibility of the initial value of the optimization algorithm, so as to ensure that the optimization design can be carried out smoothly.

In the motor design, not only logical reasoning but also analogy, association, and experience are used. The neural network has the characteristics of distributed parallelism, self-organization, self-association, fault tolerance and so on. It has strong complementarity with the expert system. An analogy method based on neural network is proposed to determine the main dimensions of the motor and the neural network-based motor design experience knowledge representation method. , And in a Y series of small three-phase asynchronous motor design success. In addition, the optimization design of Wen-Men's motor, and the fuzzification of the objective function, the satisfaction degree of the example results is greatly improved compared with the original design. It indicates that fuzzy optimization is an effective motor optimization method.

An optimization algorithm that is suitable for combinatorial optimization problems, etc., has the characteristics of a simple structure, a weak dependence on the initial point, and the ability to find the global best or the global best.

The optimal design of the motor makes full use of the SA algorithm to find the characteristics of the global optimum and the advantages of the fast GA convergence rate. In 1988, Harth et al. of the United States proposed a randomized parallel optimization algorithm, Alopex, which solves the problems of combinatorial optimization and pattern matching. This is an optimization algorithm combining heuristic and stochastic optimization. It overcomes the trapping of traditional direct methods. The defect of the local optimum, in turn, overcomes the lack of convergence of the SA algorithm. For the first time, the Alopex optimization algorithm is used in the optimization design of permanent magnet DC motors for automotive air conditioners, and satisfactory results are achieved.

In recent years, the regional elimination method has received more and more attention in the application of motor optimization design. Its basic idea is to systematically explore the entire feasible region to find the global minimum. The algorithm can avoid the repeated search near the local minimum point and all the regions that cause this minimum point, thereby increasing the chance of finding a new local minimum point in the unsearched region. When the entire area is searched, the smallest local minimum point is taken as the global minimum point, and the probability that the real global minimum point is found increases as the random point increases. The optimized design of the linear motor and the permanent magnet starter motor is respectively performed using the algorithm, and the force energy index of the linear motor and the effective material consumption of the permanent magnet machine are greatly improved, and the convergence speed is also ideal.

Particle swarm optimization (PSO) is an evolutionary computation technique (EC) that originated from the study of bird predation behavior. It was first proposed by Dr. Eberhart and Dr. Kennedy. In contrast, PSO not only has GA's global optimization ability, but also has a strong local search ability by adjusting the parameters of PS0.

Since there is no complicated operation such as individual hybridization or mutation, the parameter adjustment of PS0 becomes simple and easy, which is more suitable for computer programming. Particle swarm optimization (PS0) is a new optimization algorithm, and its application in motor optimization is not deep. According to this principle, the optimal design procedure of the linear induction motor was programmed with VB, and the bilateral non-magnetic secondary linear induction motor was optimized. As a result, the power factor and synchronization efficiency were significantly improved compared with the original scheme.

A strong attempt was made to provide this. Through the innovation of the design formula of the asynchronous motor and the proper selection of optimization variables, almost all the optimization numerical algorithms can be applied to the motor design practice without modification, and at the same time the global convergence probability (ie, the optimal The possibility of solution has also greatly improved.

In summary, the research on motor optimization design has made great progress, and most of its design quality can exceed the previous experience design. However, to fully meet the needs of the project, there are still some problems that need to be solved in the specific practice.

A complete global optimization method suitable for engineering applications has not yet been established to solve the motor optimization design problem. The highly nonlinear state of the objective function and constraint function of the motor optimization design determines that the best method of motor optimal design should adopt the nonlinear programming method based on the direct search method, but at present, there is no suitable solution method for this kind of problem; for the initial point The choices made in the current optimization process are mostly initial design schemes or expert experiences, and the flexibility is too great; comprehensive consideration should be given to the multi-objective comprehensive optimization design issues including the structure of the motor, noise and vibration, and temperature rise beyond the electromagnetic scheme. To further improve the overall optimization accuracy of the motor; the lack of standardized, general-purpose commercial optimization design software, the phenomenon of repeated studies is serious, and the level of optimization design software is low, lagging behind the development of computer software technology.

Looking at a large number of research results of motor optimization technologies at home and abroad, we can easily find that the optimization technology has been widely used in the field of motor design. Selecting the optimal method for motor design has become the key to the success of motor optimization. Therefore, the improvement of the original optimization technology and the exploration of new optimization strategies are still the focus of future work. At the same time, it should also be noted that the optimization design is only an important calculation method and cannot replace the significant progress in design, structure, craftsmanship, materials, and other aspects of the design itself. This requires that professionals should take a two-pronged approach to optimization. Designing to better master the laws of design and deepen the understanding of concepts, while continuing to adhere to basic research, the two must not be neglected, so as to achieve better results.

The Multi Pocket Safety Vest has many useful features and a number of storage solutions which make it perfect for a wide range of working environments. Such as mobile phone pocket, double pen pocket, ID pocket with window top opening and bellowed front storage pockets with velcro-fastened flap. We have over 10 colors and 5 different colour combinations style for your choose. There is also enough space on the back of the vest to print your LOGO and the style of them can be customized.

Multi Pocket Safety Vest

Multi Pocket Safety Vest,Safety Vest With Pockets,Reflective Workwear,Traffic Vest

Xinxiang Zhongke Reflective Material Co., Ltd. , https://www.safetyvest.pl