阅读说明：本技术 永磁同步电机效率最优点快速搜索方法及系统 (Method and system for rapidly searching efficiency optimal point of permanent magnet synchronous motor ) 是由 华青松 徐晓通 仙存妮 何袁生 丁茂起 李力 于 2019-09-24 设计创作，主要内容包括：本发明提供了一种永磁同步电机效率最优点快速搜索方法及系统,该方法以逆变器直流侧输入功率最小为控制目标,在给定工况下通过模糊推理在线搜索,找到使永磁同步电机输入功率最低的最优励磁电流值。最优点一旦找到,电磁转矩,转子转速以及相应的最优励磁电流值将被记录,作为神经网络的训练样本。经过在不同工况下采集的样本的广泛训练后,神经网络能够学习到转矩,转速与最优励磁电流的映射关系,这时神经网络便可取代模糊控制器,无需搜索而对任意工况下的最优励磁电流进行预测,从而使电机自适应地运行在较低能耗状态。(The invention provides a method and a system for quickly searching an efficiency optimum point of a permanent magnet synchronous motor. Once the optimal point is found, the electromagnetic torque, the rotor speed and the corresponding optimal excitation current value are recorded as training samples of the neural network. After extensive training of samples collected under different working conditions, the neural network can learn the mapping relation between the torque, the rotating speed and the optimal excitation current, and at the moment, the neural network can replace a fuzzy controller and predict the optimal excitation current under any working condition without searching, so that the motor can adaptively run in a low energy consumption state.)

1. A fast search method for an efficiency optimal point of a permanent magnet synchronous motor is characterized by comprising the following steps:

s1, under a given working condition, setting a fuzzy controller, wherein the input of the fuzzy controller is the input power of the current inverter at the direct current side, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment;

s2, finding out the optimal excitation current value with the lowest input power of the permanent magnet synchronous motor as an optimal point through fuzzy reasoning on-line search; obtaining an accurate control variable as the output of the fuzzy controller through defuzzification processing;

s3, performing per unit processing on the input and output of the fuzzy controller;

s4, recording the electromagnetic torque, the rotor speed and the optimal excitation current value corresponding to the optimal point, and taking the electromagnetic torque, the rotor speed and the optimal excitation current value as training samples of the neural network; and collecting training samples corresponding to the optimal points under different working conditions to form a training sample set, training the neural network to obtain a mapping relation between the electromagnetic torque, the rotor rotating speed and the optimal excitation current value, and replacing the neural network with the fuzzy controller to complete the control of the permanent magnet synchronous motor.

2. The method of claim 1, wherein in S1, the increment of the power value is obtained by:

wherein A, B is a constant, determined by empirical values;for the rotor speed, omega, at the previous moment_rIs the current rotor speed.

3. The method according to claim 1, wherein in S1, the excitation current increment is obtained by:

wherein C, D is a constant, determined by empirical values;

4. A method according to claim 2 or 3, characterized in that the constants are obtained from empirical values by: and according to the measured data, taking approximate values according to different working conditions, and if the set effect cannot be achieved, optimizing again until the set effect is met.

5. The method according to claim 1, wherein in S2, the fuzzy inference online search is performed by:

when the input power is reduced due to the last exciting current increment, searching in the same direction is continued; otherwise, the input power is increased and the search is carried out along the opposite direction;

the search step size and the power change are set to be in a direct proportion relation.

6. A system for fast search of efficiency optimum points of a permanent magnet synchronous motor, the system comprising at least:

the system comprises a permanent magnet synchronous motor and an optimal point prediction module;

the optimal point prediction module adopts a neural network and is used for predicting the optimal excitation current under different working conditions so as to realize the self-adaptive control of the permanent magnet synchronous motor.

7. The system of claim 6, wherein the neural network is trained by:

acquiring the input power of the direct current side of the current inverter, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment under different working conditions, and finding the optimal exciting current value with the lowest input power of the permanent magnet synchronous motor as an optimal point;

recording the electromagnetic torque, the rotor speed and the optimal excitation current value corresponding to the optimal point, taking the electromagnetic torque, the rotor speed and the optimal excitation current value as training samples of the neural network, and forming a training sample set;

training the neural network based on the sample set.

8. The system of claim 7, wherein searching for the optimal point is performed by:

setting a fuzzy controller in the system, wherein the input of the fuzzy controller is the input power of the current inverter at the direct current side, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment;

and finding the optimal excitation current value with the lowest input power of the permanent magnet synchronous motor through fuzzy reasoning on-line search, thereby determining the optimal point under a certain working condition.

9. The system of claim 8, wherein the fuzzy controller is removed from the system when the neural network training is complete.

10. The system of claim 8, wherein the fuzzy inference online search is performed by:

when the input power is reduced due to the last exciting current increment, searching in the same direction is continued;

otherwise, the input power is increased and the search is carried out along the opposite direction;

the search step size and the power change are set to be in a direct proportion relation.

Technical Field

The invention belongs to the field of control of permanent magnet synchronous motors, and particularly relates to a method and a system for quickly searching an efficiency optimum point of a permanent magnet synchronous motor.

Background

The types of the existing energy-saving control methods or control systems of the permanent magnet synchronous motor are mainly divided into two types: model (Loss Model Control, LMC) and Search (SC). The model method obtains the optimal excitation directly through calculation by establishing a minimum loss model of the permanent magnet synchronous motor, has high system response speed, needs accurate modeling, is greatly influenced by motor parameters and environmental temperature change, and has no universality for different motors. The searching method omits complicated iron core analysis, determines the optimal operating point by searching the minimum current or the minimum input power of the motor operation, does not depend on the prior information of motor parameters and models, has high optimizing precision, but has high requirements on the measurement of current, power, rotating speed and the like, and has longer algorithm convergence time. The rosenblock method, the gradient method and the gold search method are some of the search algorithms studied earlier. In recent years, the development of intelligent control technology has provided powerful tools for the study of motor systems characterized by multivariable, strong coupling, nonlinearity, large hysteresis, time-varying, etc. Methods such as sliding mode variable structures, genetic algorithms, fuzzy control, artificial neural networks and the like are introduced into the field of motor control in a dispute. However, at present, conventional methods such as fuzzy control and neural network do not have a set of systematic method for users to use, and are mostly specific type setting modes based on specific working conditions or control targets, so that the method has no good expandability, and the general algorithm process is complex and the resource consumption cost is high.

Based on the above existing problems, a control system or method that can meet the requirements of the market on efficiency, resource consumption and control accuracy in a balanced manner is needed to improve the operation efficiency of the motor, realize self-adaptive energy-saving operation of the motor, and can be well applied industrially.

Disclosure of Invention

Aiming at the defects of the prior art, the invention adopts fuzzy reasoning online search and combines a neural network to realize the quick search and the self-adaptive energy-saving operation of the motor efficiency optimum point, and provides a quick search method of the permanent magnet synchronous motor efficiency optimum point and a system in industrial application thereof, so as to realize the purposes of improving the operation efficiency of the motor and realizing the self-adaptive energy-saving operation of the motor. Specifically, the invention provides the following technical scheme:

on one hand, the invention provides a method for quickly searching an efficiency optimal point of a permanent magnet synchronous motor, which comprises the following steps:

s1, under a given working condition, setting a fuzzy controller, wherein the input of the fuzzy controller is the input power of the current inverter at the direct current side, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment;

s2, finding out the optimal excitation current value with the lowest input power of the permanent magnet synchronous motor as an optimal point through fuzzy reasoning on-line search; obtaining an accurate control variable as the output of the fuzzy controller through defuzzification processing;

s3, performing per unit processing on the input and output of the fuzzy controller;

s4, recording the electromagnetic torque, the rotor speed and the optimal excitation current value corresponding to the optimal point, and taking the electromagnetic torque, the rotor speed and the optimal excitation current value as training samples of the neural network; and collecting training samples corresponding to the optimal points under different working conditions to form a training sample set, training the neural network to obtain a mapping relation between the electromagnetic torque, the rotor rotating speed and the optimal excitation current value, and replacing the neural network with the fuzzy controller to complete the control of the permanent magnet synchronous motor.

Preferably, in S1, the increment of the power value is obtained by:

wherein A, B is a constant, determined by empirical values;

for the rotor speed, omega, at the previous moment_rIs the current rotor speed.

Preferably, in S1, the excitation current increment is obtained by:

wherein C, D is a constant, determined by empirical values;

for the rotor speed, omega, at the previous moment_rIs the current rotor speed; t is_eIn order to be able to set the electromagnetic torque,

the rated electromagnetic torque at the last moment.

Preferably, the constant is obtained from empirical values by: and according to the measured data, taking approximate values according to different working conditions, and if the set effect cannot be achieved, optimizing again until the set effect is met.

Preferably, in S2, the fuzzy inference online search is performed by:

when the input power is reduced due to the last exciting current increment, searching in the same direction is continued; otherwise, the input power is increased and the search is carried out along the opposite direction;

the search step size and the power change are set to be in a direct proportion relation.

On the other hand, the invention also provides a system for rapidly searching the efficiency optimum point of the permanent magnet synchronous motor, which at least comprises the following components:

the system comprises a permanent magnet synchronous motor and an optimal point prediction module;

the optimal point prediction module adopts a neural network and is used for predicting the optimal excitation current under different working conditions so as to realize the self-adaptive control of the permanent magnet synchronous motor.

Preferably, the neural network is trained by:

acquiring the input power of the direct current side of the current inverter, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment under different working conditions, and finding the optimal exciting current value with the lowest input power of the permanent magnet synchronous motor as an optimal point;

recording the electromagnetic torque, the rotor speed and the optimal excitation current value corresponding to the optimal point, taking the electromagnetic torque, the rotor speed and the optimal excitation current value as training samples of the neural network, and forming a training sample set;

training the neural network based on the sample set.

Preferably, searching for the optimal point is performed by:

setting a fuzzy controller in the system, wherein the input of the fuzzy controller is the input power of the current inverter at the direct current side, the increment of the power value at the previous sampling moment and the increment of the exciting current at the previous sampling moment;

and finding the optimal excitation current value with the lowest input power of the permanent magnet synchronous motor through fuzzy reasoning on-line search, thereby determining the optimal point under a certain working condition. When the fuzzy controller searches for the optimal point under a plurality of working conditions, a sufficient number of training samples are formed, and then the neural network can be trained. Of course, the specific number of training samples needs to be adjusted based on the accuracy requirement and the structural complexity of the neural network, which can be determined by those skilled in the art based on the disclosure of the present invention in combination with the common general knowledge in the art, and will not be described herein again.

Preferably, the fuzzy controller is removed from the system when the neural network training is completed. Of course, the system may still maintain the fuzzy controller for use in situations such as motor type replacement, neural network model adjustment, etc., where it is necessary to perform alternative control or to re-collect the optimal point sample data.

Preferably, the fuzzy inference online search is performed by:

when the input power is reduced due to the last exciting current increment, searching in the same direction is continued; otherwise, the input power is increased and the search is carried out along the opposite direction;

the search step size and the power change are set to be in a direct proportion relation.

Compared with the prior art, the rapid search algorithm adopted by the invention improves the operation efficiency of the motor and realizes the self-adaptive energy-saving operation of the motor.

Drawings

FIG. 1 is a block diagram of a fuzzy inference online search controller according to an embodiment of the present invention;

fig. 2 is a diagram illustrating an operation state of a permanent magnet synchronous motor according to an embodiment of the present invention;

FIG. 3a is a rotor speed plot of the dynamic response of a permanent magnet synchronous motor in accordance with an embodiment of the present invention;

FIG. 3b is an electromagnetic torque diagram of the dynamic response of a permanent magnet synchronous machine of an embodiment of the present invention;

fig. 3c is a diagram of the dc side input power of the dynamic response of the permanent magnet synchronous motor according to the embodiment of the present invention.

DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention.

The technical solution of the present invention will be further described in detail with reference to the accompanying drawings and embodiments.

The invention provides a method for rapidly searching an efficiency optimal point of a permanent magnet synchronous motor and a corresponding system in order to improve the operation efficiency of the motor and realize self-adaptive energy-saving operation of the motor. In a specific embodiment, the method may be implemented as follows:

step 1, the whole optimization method takes the minimum input power at the direct current side of the inverter as a control target, and finds the optimal excitation current value which enables the minimum input power of the permanent magnet synchronous motor through fuzzy reasoning on-line search under a given working condition (for example, under the condition of certain torque and rotor rotating speed).

In a preferred embodiment, the fuzzy controller may take a dual input single output configuration, as shown in FIG. 1. The input of the fuzzy system is the increment of the power value of the current inverter direct-current side input power and the previous sampling moment, and the increment of the excitation current at the previous sampling moment, fuzzy output is obtained through fuzzy reasoning and fuzzy decision, and the accurate control variable is obtained through defuzzification.

Step 2, in order to enable the fuzzy controller to be applied to motors with different powers and different parameters, per unit processing is needed, and input and output are uniformly set to a normalized universe of discourse (-1, 1). Wherein the power increment and the exciting current increment are obtained according to the following formula:

a, B, C, D is constant, which is determined empirically, and the above constant is determined by empirical value, and it can be approximated by previous measured data according to different working conditions, and if the predetermined effect is not achieved, then it is optimized, and when the predetermined effect is achieved, the corresponding constant is determined.

The rotor speed at the previous moment,

Rated electromagnetic torque, T, for the preceding moment_eThe rated electromagnetic torque is obtained by an estimation module; omega_rIs the current rotor speed.

The proportional factor is a function of the rotating speed and the torque estimated value, so that the exciting current output by the fuzzy control has a self-adaptive function on the torque and the rotating speed, and the convergence speed is optimized.

Step 3, when the last exciting current increment

Input power deviation Δ P when input power is caused to decrease_d(Pu) is less than 0, and searching continues in the same direction; otherwise, the input power rises, the input power deviation Δ P_d(Pu) is greater than 0, and the search is in the opposite direction, the step size of the search being proportional to the power change.

And 4, once the optimal point is found, recording the electromagnetic torque, the rotor rotating speed and the corresponding optimal excitation current value as a training sample of the neural network. After extensive training of samples collected under different working conditions (different starting torque torques, different input voltages and the like), the neural network can learn the mapping relation between the torque, the rotating speed and the optimal excitation current, and at the moment, the neural network can replace a fuzzy controller and predict the optimal excitation current under any working condition without searching, so that the motor can adaptively operate in a low energy consumption state, as shown in fig. 2.

The following examples are provided to further illustrate the technical act of the present invention: based on MATLAB/Simulink environment, the rapid search algorithm is subjected to simulation verification, a permanent magnet synchronous motor with the rated power of 2Kw, the rated voltage of 380V and the frequency of 50Hz is selected as a simulation motor, and the sampling period is T_s2e-6s, the search step size is 0.2 s. As shown in FIG. 3, 0s starts the motor, load torque is 4.5 N.m (0.3pu), and rotational speed is 1100r/min (0.8 pu). The first 1.5s is the starting time of the motor, and then the motor enters into steady-state operation. The fast search algorithm is added at 1.5 s. Electromagnetic torque T due to implementation of torque current feed-forward compensation_eAnd the rotor speed n does not change significantly, and the output power P_dFrom 830W down to 730W. The feasibility of the rapid search algorithm is verified, the operation efficiency of the motor can be improved, and the self-adaptive energy-saving operation of the motor is realized.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.