Intelligent torque method for optimizing adaptive fuzzy PID based on DWOA algorithm

Abstract

The invention relates to an intelligent torque method for optimizing self-adaptive fuzzy PID based on a DWOA algorithm, and the method comprises the steps: adjusting the parameters Kp, Ki and Kd of a PID controller, enabling the output rotating speed of a motor to be close to a set value, constructing the self-adaptive fuzzy PID, employing a square deviation integral ISE as a fitness function, employing a surrounding prey, a bubble net attack and a searching prey to optimize a whale algorithm, and carrying out the optimization of the self-adaptive fuzzy PID. The whale individual is set, the whale position is updated and discretized, after PID parameters are updated each time, the needed control torque u (t) is calculated through new parameters, the finally obtained optimal PID parameter combination is applied to an actual controller, and the needed torque is output; according to the method, the discrete optimization problem of the intelligent torque is solved through the discrete whale optimization algorithm, the torque accurate control in the dynamically changing environment is realized in combination with the self-adaptive fuzzy PID, and the accurate and stable constant-speed rotation control is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of torque control, and in particular to an intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID. Background Art

[0002] The high-voltage cabinet is a closed structure. The operation and maintenance personnel cannot timely and effectively judge the safe and stable operation of the equipment before operating the switch. At the same time, due to the closed structure, the energized state of the equipment can only be judged based on electrical or mechanical indications, and direct electrical testing cannot be performed. When the electrical or mechanical indications are abnormal or damaged, it poses a threat to the personal safety of the operator; when detecting and eliminating hidden faults (equipment inter-turn short circuit, intermittent grounding), directly operating the switch has certain safety hazards. The present invention can automatically adjust the torque applied to the cabinet door under different operating conditions through adaptive torque control to ensure that the operation of opening the cabinet door will not cause additional risks due to the energized state or abnormal indication. Compared with traditional manual or mechanical operation, the electric method is more controllable and can better avoid safety threats caused by misoperation. The built-in adaptive control system and constant speed stabilization technology can improve the reliability of equipment operation and reduce the failure of cabinet door opening due to operating device failure or signal transmission problems. Summary of the invention

[0003] The present invention provides an intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID, and aims at an external electric remote control operation device of a non-automated medium voltage cabinet to make its output torque adaptive, so that the cabinet door can be opened at a constant speed.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] An intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID, comprising the following steps:

[0006] S1. Define the problem: By adjusting the K of the PID controller p , K i , K d Parameters, so that the motor output speed is close to the set value, the input is the error e(t) and the error change rate The error is the difference between the actual speed and the desired speed, and the output is the control signal u(t), which is used to adjust the motor torque;

[0007] S2. Constructing adaptive fuzzy PID

[0008] (1) Construct a fuzzy rule base based on e(t) and Determine the increment ΔK of PID parameters p , ΔK i , ΔK d ;

[0009] (2) Design the membership function to map e(t) and to the corresponding fuzzy set;

[0010] (3) Use the centroid method to perform defuzzification and obtain the specific PID parameter adjustment amount;

[0011] S3, DWOA algorithm initialization

[0012] (1) Population initialization: N individuals are randomly generated, each of which represents a set of PID parameter combinations (K p,i , K i,i , K d,i ), where i = 1, 2, ..., N;

[0013] (2) Position vector: The position vector of each individual is X i =(K p,i ,K i,i ,K d,i );

[0014] S4, fitness function definition: squared deviation integral ISE is used as the fitness function;

[0015] S5, DWOA algorithm execution

[0016] (1) Surrounding prey: When humpback whales find the location of their prey during hunting, they gradually surround it. The algorithm simulates the behavior of other individuals in the whale population moving toward the optimal individual position.

[0017] (2) Bubble net attack: Construct a spiral position update to describe the predation behavior, that is, each whale individual adopts a spiral movement to approach the current leading whale individual;

[0018] (3) Searching for prey: Based on mathematical modeling of the shrinking and encircling predation strategy, individual whales approach the currently known optimal solution;

[0019] S6, combined with intelligent torque: After each PID parameter update, the required control torque u(t) is calculated using the new parameters:

[0020]

[0021] By adjusting the torque u(t), the motor can respond in time and maintain a stable speed;

[0022] S7, iteration and convergence: Repeat the above process until the stop condition is met, and apply the best PID parameter combination finally obtained to the actual controller to output the required torque.

[0023] Furthermore, the encirclement of the prey adopts shrinking encirclement, and the position update formula is as follows:

[0024] X(t+1)=X * (t)-A·D (2)

[0025] D=|C·X * (t)-X(t)| (3)

[0026] Among them, vectors A and C are coefficient vectors that control the search behavior, t is the current iteration number, and X * (t) refers to the optimal solution position found so far, and X(t) represents the current position state. The specific calculation method of vectors A and C is as follows:

[0027]

[0028] Among them, a will decrease linearly from 2 to 0 with the increase of iterations; r is a random variable in |0,1].

[0029] Furthermore, the bubble net attack imitates the humpback whale's bubble hunting method, that is, the bubble net attack, and the mathematical model is as follows:

[0030]

[0031] It is used when |A|<1, where p represents a random value in the interval [0, 1], the constant b is the morphological control parameter of the spiral path, l is a random value in the interval [-1, 1], and D′=|X * (t)-X(t)|,

[0032] Furthermore, the prey search mathematical model is as follows:

[0033]

[0034] D=|C·X rand -X(t)| (7)

[0035] Used when |A|≥1, where X rand is the randomly selected whale position, is the distance vector between the whale and its prey.

[0036] Furthermore, the DWOA algorithm performs setting of individual whales and discretizes the whale position update, wherein the setting of individual whales is that an individual whale contains multiple elements, each element represents an activity, and the index value of each element represents the order of activity execution, and the process problem of multiple bending points is solved by using the element arrangement method;

[0037] The above discretization of whale position update sets the current best whale individual as X *, the current whale individual is X j , then the current whale individual is X j+1 The update formula is as follows:

[0038]

[0039] Among them, C 1 is the first crossover method for two whale individuals. 2 The second crossover method for performing a crossover operation on two whale individuals, where p represents a random number and takes a value of [0, 1].

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The discrete whale optimization algorithm is used to solve the discrete optimization problem of intelligent torque, and the adaptive fuzzy PID is combined to achieve precise torque control in a dynamically changing environment, and achieve precise and stable constant speed rotation control. The adaptive torque control can automatically adjust the torque under different conditions to ensure smooth opening and closing of the cabinet door. Even if the electrical or mechanical indications are abnormal, it can be operated safely. It can ensure that the cabinet door opens and closes smoothly while maintaining a constant speed, reducing the risk of equipment damage caused by opening the door too fast or too slow. The constant speed and stable opening and closing operation can significantly reduce the mechanical vibration during the opening and closing operation, and reduce the risk of flashover grounding or short circuit caused by vibration. DETAILED DESCRIPTION

[0042] The specific embodiments of the present invention are further described below:

[0043] The present invention provides an intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID, comprising the following steps:

[0044] S1. Define the problem: By adjusting the K of the PID controller p , K i , K d Parameters, so that the motor output speed is close to the set value, the input is the error e(t) and the error change rate The error is the difference between the actual speed and the desired speed, and the output is the control signal u(t), which is used to adjust the motor torque.

[0045] S2. Constructing adaptive fuzzy PID: Adaptive fuzzy PID control combines the ideas of fuzzy control and adaptive control on the basis of conventional PID algorithm. Compared with conventional PID algorithm, it is mainly used to improve the stability and robustness of the system. It is suitable for complex time-varying nonlinear systems. Under the premise of not changing the control effect of the original PID algorithm, the adaptive fuzzy PID controller calculates the control output through the deviation and deviation change rate of turbidity, and according to the actual operation of the system, the parameters of the PID controller are adaptively adjusted in real time to maintain the stability and performance of the system and achieve precise control in a dynamically changing environment.

[0046] Adaptive fuzzy PID control uses the fuzzy control rules established in the fuzzy controller to incrementally adjust the three parameters of the PID controller. According to the fuzzy relationship between the deviation e between the set turbidity and the measured turbidity and the deviation change rate ec and the fuzzy quantity of the three PID parameter increments in the fuzzy controller, the turbidity deviation e and the deviation change rate ec are updated in real time during the operation of the adaptive fuzzy PID control system. The increments of the three PID parameters are adjusted online according to the set fuzzy control rules, and the K of the PID controller is adaptively adjusted. p , K i , K d The formula is:

[0047] K p =K p0 +λ p ΔK p (9)

[0048] K i =K i0 +λ i ΔK i (10)

[0049] K d =K d0 +λ d ΔK d (11)

[0050] In the formula, K p0 , K i0 , K d0 is the initial parameter pre-adjusted by the PID controller, ΔK p , ΔK i , ΔK d The outputs of the fuzzy controller correspond to K p , K i , K d The amount of blurring of the variation, λ p , i , d They are ΔK p, ΔK i , ΔK d The proportional factor, when the deviation e between the set turbidity and the measured turbidity and the deviation change rate ec change, ΔK p , ΔK i , ΔK d The corresponding defuzzified value will also change accordingly, so the K of the PID controller p , K i , K d The parameters of the PID controller will be adaptively adjusted according to the real-time operating status of the control system to adapt to the dynamic changes of the system and external interference, thus improving the adaptive ability of the control system, so that it has a better control effect in all control processes, as follows:

[0051] (1) Construct a fuzzy rule base based on e(t) and Determine the increment ΔK of PID parameters p , ΔK i , ΔK d ;

[0052] (2) Design the membership function to map e(t) and to the corresponding fuzzy set;

[0053] (3) The centroid method is used for defuzzification to obtain the specific PID parameter adjustment amount.

[0054] S3, DWOA algorithm initialization

[0055] (1) Population initialization: N individuals are randomly generated, each of which represents a set of PID parameter combinations (K p,i , K i,i , K d,i ), where i = 1, 2, ..., N;

[0056] (2) Position vector: The position vector of each individual is X i =(K p,i ,K i,i ,K d,i );

[0057] S4, fitness function definition: squared deviation integral ISE is used as the fitness function, the function is as follows:

[0058]

[0059] Where J is the integrated square error, which measures the square of the integral value of the system response error over a given period of time (from time 0 to time T) multiplied by time. Here the error is defined as the difference between the expected output and the actual output.

[0060] T is the time upper limit of the square deviation integration;

[0061] e(t) is the error function, which represents the deviation between the system response and the actual response;

[0062] dt is the differential symbol, which means time differential.

[0063] S5, DWOA algorithm execution

[0064] Whale Optimization Algorithm (WOA) is an intelligent optimization algorithm that simulates the predation behavior of humpback whale groups. Whale individuals represent a solution to the problem to be solved. Whale individuals update their positions through three methods: shrinking encirclement, circling predation, and random predation. The mathematical models of the three update methods are as follows:

[0065] (1) Surrounding prey: When humpback whales find the location of their prey during hunting, they gradually surround it. The WOA algorithm simulates the behavior of other individuals in the whale population moving toward the optimal individual position. The position update formula is as follows:

[0066] X(t+1)=X * (t)-A·D (13)

[0067] D=|C·X * (t)-X(t)| (14)

[0068] Among them, vectors A and C are coefficient vectors that control the search behavior, t is the current iteration number, and X * (t) refers to the optimal solution position found so far, and X(t) represents the current position state. The specific calculation method of vectors A and C is as follows:

[0069]

[0070] Among them, a will decrease linearly from 2 to 0 as the number of iterations increases; r is a random variable in [0,1]

[0071] (2) Bubble net attack: In order to imitate the unique bubble hunting method of humpback whales, namely bubble net attack, a spiral position update was constructed to describe this predation behavior, that is, each whale individual adopts a spiral movement method with the aim of approaching the current leading whale individual. The formula is as follows:

[0072] X(t+1)=D′·e bl ·cos(2πl)+X * (t) (16)

[0073] D′=|X * (t)-X(t)| (17)

[0074] Among them, the constant b is the morphological control parameter of the spiral path, and l is a random value in the interval [-1, 1]. The idea of ​​this algorithm comes from the unique hunting method of humpback whales, especially their behavior in the process of chasing prey: using a spiral path to approach prey. The adoption of this method aims to simulate the flexibility and efficiency of humpback whales when hunting in the natural environment, so as to fully explore and utilize the potential solution space when solving optimization problems. Its mathematical model is as follows:

[0075]

[0076] Here, p represents a random value in the interval [0, 1].

[0077] (3) Searching for prey: For the mathematical modeling of the shrinking and encircling predation strategy, the value of coefficient A is designed to vary within the range of [-1, 1], thereby guiding individual whales to approach the currently known optimal solution (i.e., the current best individual whale). However, when the value of A exceeds this range, the algorithm introduces a variant strategy, that is, it no longer simply approaches the optimal solution, but randomly selects an individual within the whale group to approach. This mechanism embodies the concept of search and predation. Its core idea is that although the movement direction of individual whales may temporarily deviate from the current target prey, this deviation helps to expand the exploration range of the entire whale group, thereby enhancing the global search ability of the algorithm; the mathematical model of search and predation behavior is as follows:

[0078]

[0079] D=|C·X rand -X(t)| (20)

[0080] Among them, X rand is the randomly selected whale location;

[0081] Since the classic whale optimization algorithm can only handle continuous problems, and the intelligent torque is a discrete optimization problem, not only does it need a new concept, but in the classic whale optimization algorithm, individual whales are represented by coordinates of different dimensions, and the new position coordinates are updated by the corresponding update method of capturing prey. Therefore, the update of individual whales and search positions has also been modified, including the following:

[0082] (1) Whale individual setting: Use an element arrangement method to solve the process problem of multiple bending points; a whale individual contains multiple elements, each element represents an activity, and the index value of each element indicates the order of activity execution;

[0083] (2) Discretization of whale position update: Assume that the current best whale individual is X* , the current whale individual is X j , then the current whale individual is X j+1 The update formula is as follows:

[0084]

[0085] Among them, C 1 is the first crossover method for two whale individuals. 2 The second crossover method for performing a crossover operation on two whale individuals, where p represents a random number and takes a value of [0, 1].

[0086] S6, combined with intelligent torque: After each PID parameter update, the required control torque u(t) is calculated using the new parameters:

[0087]

[0088] By adjusting the torque u(t), the motor can respond in time and maintain a stable speed.

[0089] S7, iteration and convergence: Repeat the above process until the stopping condition is met. The stopping condition is the maximum number of iterations or a satisfactory performance indicator is achieved. The best PID parameter combination finally obtained is applied to the actual controller to achieve accurate and stable constant speed rotation control.

[0090] The above embodiments are implemented based on the technical solution of the present invention, and detailed implementation methods and specific operation processes are given, but the protection scope of the present invention is not limited to the above embodiments. The methods used in the above embodiments are conventional methods unless otherwise specified.

Claims

1. An intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID, characterized in that: The steps include: S1. Define the problem: By adjusting the K of the PID controller p , K i , K d Parameters, so that the motor output speed is close to the set value, the input is the error e(t) and the error change rate The error is the difference between the actual speed and the desired speed, and the output is the control signal u(t), which is used to adjust the motor torque; S2. Constructing adaptive fuzzy PID (1) Construct a fuzzy rule base based on e(t) and Determine the increment ΔK of PID parameters p , ΔK i , ΔK d ; (2) Design the membership function to map e(t) and to the corresponding fuzzy set; (3) Use the centroid method to perform defuzzification and obtain the specific PID parameter adjustment amount; S3, DWOA algorithm initialization (1) Population initialization: N individuals are randomly generated, each of which represents a set of PID parameter combinations (K p,i , K i,i , K d,i ), where i = 1, 2, ..., N; (2) Position vector: The position vector of each individual is X i =(K p,i ,K i,i ,K d,i ); S4, fitness function definition: squared deviation integral ISE is used as the fitness function; S5, DWOA algorithm execution (1) Surrounding prey: When humpback whales find the location of their prey during hunting, they gradually surround it. The algorithm simulates the behavior of other individuals in the whale population moving toward the optimal individual position. (2) Bubble net attack: Construct a spiral position update to describe the predation behavior, that is, each whale individual adopts a spiral movement to approach the current leading whale individual; (3) Searching for prey: Based on mathematical modeling of the shrinking and encircling predation strategy, individual whales approach the currently known optimal solution; S6, combined with intelligent torque: After each PID parameter update, the required control torque u(t) is calculated using the new parameters: By adjusting the torque u(t), the motor can respond in time and maintain a stable speed; S7, iteration and convergence: Repeat the above process until the stop condition is met, and apply the best PID parameter combination finally obtained to the actual controller to output the required torque.

2. According to claim 1, the intelligent torque method based on DWOA algorithm to optimize the adaptive fuzzy PID is characterized in that: The encirclement of the prey adopts shrinkage encirclement, and the position update formula is as follows: X(t+1)=X * (t)-A·D (2) D=|C·X * (t)-X(t)| (3) Among them, vectors A and C are coefficient vectors that control the search behavior, t is the current iteration number, and X * (t) refers to the optimal solution position found so far, and X(t) represents the current position state. The specific calculation method of vectors A and C is as follows: Among them, a will decrease linearly from 2 to 0 as the number of iterations increases; r is a random variable in [0,1].

3. The intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID according to claim 1 is characterized in that: The bubble net attack imitates the humpback whale's bubble hunting method, that is, the bubble net attack, and the mathematical model is as follows: It is used when |A|<1, where p represents a random value in the interval [0, 1], the constant b is the morphological control parameter of the spiral path, l is a random value in the interval [-1, 1], and D′=|X * (t)-X(t)|.

4. According to claim 1, the intelligent torque method based on DWOA algorithm to optimize the adaptive fuzzy PID is characterized in that: The prey search mathematical model is as follows: D=|C·X rand -X(t)| (7) Used when |A|≥1, where X rand is the randomly selected whale position, is the distance vector between the whale and its prey.

5. The intelligent torque method based on DWOA algorithm to optimize adaptive fuzzy PID according to claim 1 is characterized in that: The DWOA algorithm performs the setting of individual whales and discretizes the whale position update. The setting of individual whales is that a whale individual contains multiple elements, each element represents an activity, and the index value of each element represents the order of activity execution. The process problem of multiple bending points is solved by using the element arrangement method; The above discretization of whale position update sets the current best whale individual as X * , the current whale individual is X j , then the current whale individual is X j+1 The update formula is as follows: Among them, C1 is the first crossover method for performing a crossover operation on two whale individuals, and C2 is the second crossover method for performing a crossover operation on two whale individuals, where p represents a random number, and its value is [0, 1].

Images