Frequency converter control method, system and device based on fuzzy neural network and medium

Abstract

The invention discloses a frequency converter control method, system and device based on a fuzzy neural network and a medium, and the method comprises the steps: obtaining sensor sensing data in real time, inputting the sensor sensing data into the fuzzy neural network, and obtaining a frequency converter output value through the fuzzy neural network and a PID controller in sequence; inputting the output value of the frequency converter into the fuzzy neural network in a back propagation manner, and optimizing the control parameter value of the PID controller according to the output value of the frequency converter to obtain a PID optimized parameter value; utilizing a PID controller to obtain a frequency converter adjusting signal according to the PID optimization parameter value; and controlling the output end equipment according to the frequency converter adjusting signal. The self-adaptive control of the frequency converter can be effectively realized, so that the output end equipment arranged at the output end of the frequency converter is quickly stabilized at a set value, the operation control effect on the output end equipment is more excellent, the energy consumption is lower, and the equipment can be effectively ensured to operate with high precision and high efficiency under complex working conditions.

Description

Technical Field

[0001] The present invention relates to the field of frequency converters, and particularly to a frequency converter control method, system, device and medium based on a fuzzy neural network. Background Art

[0002] In recent years, frequency converters have been widely used in many fields such as industry, transportation, energy, and household appliances. For example, motors in the industrial field, electric vehicles in the transportation field, photovoltaic power generation systems in the energy field, air conditioners and range hoods in the household appliance field, etc. In these fields, through the control of the frequency converter, purposes such as energy saving, improvement of production efficiency, improvement of equipment performance, and improvement of system stability can be achieved.

[0003] Currently, frequency converters usually adopt methods such as V / F control method (i.e., voltage-frequency ratio control method), direct torque control method, etc. to achieve control. In these control technologies, defects such as large energy consumption, over-frequency, too long lag time, and large error often occur.

[0004] Therefore, it is urgent to develop a new frequency converter control technology to effectively solve the above defects. Summary of the Invention

[0005] In view of this, the present invention provides a frequency converter control method, system, device and medium based on a fuzzy neural network to solve the problems of large energy consumption, over-frequency, too long lag time, and large error in the existing frequency converter control technology.

[0006] The present invention provides a frequency converter control method based on a fuzzy neural network, and the method includes:

[0007] Obtain sensor perception data in real time, and input the sensor perception data into the fuzzy neural network, and successively use the fuzzy neural network and the PID controller to obtain the output value of the frequency converter; wherein, the PID controller is arranged between the fuzzy neural network and the frequency converter;

[0008] Backpropagate the output value of the frequency converter into the fuzzy neural network, and optimize the control parameter value of the PID controller according to the output value of the frequency converter to obtain the PID optimization parameter value;

[0009] Use the PID controller to obtain a frequency converter adjustment signal according to the PID optimization parameter value;

[0010] Control the output device of the frequency converter according to the frequency converter adjustment signal.

[0011] Optionally, the fuzzy neural network includes an input layer, an inference layer and an output layer arranged in sequence;

[0012] Optimizing the control parameter value of the PID controller according to the output value of the frequency converter to obtain the PID optimization parameter value, including:

[0013] Input the output value of the frequency converter into the input layer as the input variable of the input layer; and use the input layer to calculate the deviation function according to the input variable of the input layer;

[0014] Input the deviation function into the inference layer, and calculate the input variable of the inference layer according to the deviation function; and use the inference layer to calculate the output function of the inference layer according to the input variable of the inference layer;

[0015] Input the output function of the inference layer into the output layer, and calculate the input variable of the output layer according to the output function of the inference layer; and use the output layer to calculate the PID optimization parameter value according to the input variable of the output layer.

[0016] Optionally, the specific calculation formula of the deviation function is:

[0017] e(t) = Er(t) - Eout(t);

[0018] Wherein, e(t) is the output function of the input layer at the t-th sampling period, representing the deviation function at the t-th sampling period; Eout(t) is the input variable of the input layer at the t-th sampling period, representing the output value of the frequency converter at the t-th sampling period; Er(t) is the set value corresponding to the output end of the frequency converter at the t-th sampling period.

[0019] Optionally, there are multiple first input parameters corresponding to the input variable of the inference layer at the t-th sampling period. Correspondingly, there are also multiple first output parameters corresponding to the output function of the inference layer at the t-th sampling period, and the number of the first output parameters is the same as the number of the first input parameters;

[0020] The specific formula for calculating the input variable of the inference layer according to the deviation function at the t-th sampling period is:

[0021]

[0022] Wherein, is the x-th first input parameter in the input variable of the inference layer at the t-th sampling period, n1 is the first iteration number corresponding to the input variable of the inference layer at the t-th sampling period, e k (t) is the value of the deviation function at the k-th iteration at the t-th sampling period, is the inference layer connection weight of the x-th first input parameter at the k-th iteration at the t-th sampling period;

[0023] The specific formula for calculating the output function of the inference layer based on the input variables of the inference layer in the t-th sampling period is as follows:

[0024]

[0025] where, is the x-th first output parameter in the output function of the inference layer in the t-th sampling period, and tanh(·) is the hyperbolic tangent function.

[0026] Optionally, there are multiple second input parameters corresponding to the input variables of the output layer in the t-th sampling period. Correspondingly, there are also multiple second output parameters corresponding to the output function of the output layer in the t-th sampling period, and the number of the second output parameters is the same as the number of the second input parameters;

[0027] The specific formula for calculating the input variables of the output layer based on the output function of the inference layer in the t-th sampling period is as follows:

[0028]

[0029] where, is the y-th second input parameter in the input variables of the output layer in the t-th sampling period, n2 is the second iteration number corresponding to the input variables of the output layer in the t-th sampling period, is the y-th first output parameter in the output function of the inference layer in the t-th sampling period, is the output layer connection weight of the y-th second input parameter in the j-th iteration in the t-th sampling period;

[0030] The specific formula for calculating the output function of the output layer based on the input variables of the output layer in the t-th sampling period is as follows:

[0031]

[0032] where, is the y-th second output parameter in the output function of the output layer in the t-th sampling period, representing the y-th optimization parameter value in the PID optimization parameter values in the t-th sampling period; tanh(·) is the hyperbolic tangent function.

[0033] Optionally, the PID optimization parameter values include an optimization proportional coefficient, an optimization integral coefficient, and an optimization differential coefficient;

[0034] Then the optimization proportional coefficient K p (t) is specifically f1 (3) (t), and the optimization integral coefficient K i (t) is specifically The optimized differential coefficient K at the t-th sampling period d (t) is specifically

[0035] Optionally, the frequency converter adjustment signal is specifically the equipment load corresponding to the output terminal device of the frequency converter;

[0036] Then, according to the PID optimization parameter value at the t-th sampling period, the specific formula for calculating the frequency converter adjustment signal at the t-th sampling period is:

[0037] U(t) = K p (t)·e(t) + K i (t)·∑e(t) + K d (t)·[e(t) - e(t - 1)];

[0038] Wherein, U(t) is the equipment load at the t-th sampling period, and e(t - 1) is the deviation function at the (t - 1)-th sampling period.

[0039] Optionally, the output terminal device of the frequency converter has an actual collected load;

[0040] Controlling the output terminal device of the frequency converter according to the frequency converter adjustment signal includes:

[0041] Comparing the equipment load with the actual collected load;

[0042] When the equipment load is greater than the actual collected load, using the frequency converter to perform a loading operation on the output terminal device; when the equipment load is less than the actual collected load, using the frequency converter to perform a unloading operation on the output terminal device; when the equipment load is equal to the actual collected load, maintaining the current operation of the frequency converter.

[0043] Optionally, inputting the sensor sensing data into a fuzzy neural network, and sequentially using the fuzzy neural network and a PID controller to obtain a frequency converter output value, including:

[0044] Performing a fuzzification process on the sensor sensing data to obtain a fuzzy linguistic variable;

[0045] Inputting the fuzzy linguistic variable into the fuzzy neural network to obtain fuzzy rules;

[0046] Performing a fuzzy inference operation according to the fuzzy rules to obtain a fuzzy adjustment amount corresponding to the PID controller;

[0047] Defuzzifying the fuzzy adjustment amount to obtain a control parameter value of the PID controller;

[0048] Input the control parameter value of the PID controller into the PID controller to obtain a frequency converter control signal;

[0049] Input the frequency converter control signal into the frequency converter, and collect the output value of the frequency converter.

[0050] In addition, the present invention also provides a frequency converter control system based on a fuzzy neural network, which is applied to the aforementioned frequency converter control method based on a fuzzy neural network. The system includes:

[0051] A forward propagation module, configured to obtain sensor perception data in real time, input the sensor perception data into the fuzzy neural network, and successively use the fuzzy neural network and the PID controller to obtain the output value of the frequency converter; wherein, the PID controller is arranged between the fuzzy neural network and the frequency converter;

[0052] A feedback adjustment module, configured to back-propagate and input the output value of the frequency converter into the fuzzy neural network, optimize the control parameter value of the PID controller according to the output value of the frequency converter, and obtain a PID optimization parameter value;

[0053] A PID optimization module, configured to use the PID controller to obtain a frequency converter adjustment signal according to the PID optimization parameter value;

[0054] A frequency converter control module, configured to control the output device of the frequency converter according to the frequency converter adjustment signal.

[0055] In addition, the present invention also provides a frequency converter control device based on a fuzzy neural network, including a processor, a memory, and a computer program stored in the memory and operable on the processor. When the computer program runs, it implements the method steps in the aforementioned frequency converter control method based on a fuzzy neural network.

[0056] In addition, the present invention also provides a computer storage medium, which includes: at least one instruction, and when the instruction is executed by a computer, it implements the method steps in the aforementioned frequency converter control method based on a fuzzy neural network.

[0057] Advantages of the present invention: First, the output value of the frequency converter is obtained through the fuzzy neural network and the PID controller to achieve the forward propagation of the frequency converter control. Then, the obtained output value of the frequency converter is fed back and input into the fuzzy neural network. Using the backpropagation of the fuzzy neural network, based on the actually collected output value and set value of the frequency converter and the error between the two, the parameters are adjusted to optimize the control parameter value of the PID controller and obtain the optimized PID parameter value. The optimized PID parameter value is then transmitted to the PID controller to realize the self-adjustment of the PID controller parameters, enabling the PID controller to have the ability of adaptive adjustment control. Under the control of the adaptively adjusted PID controller, a frequency converter adjustment signal is obtained. Finally, the output device at the output end of the frequency converter is controlled by using this frequency converter adjustment signal, which can effectively realize the adaptive control of the frequency converter, enabling the output device set at the output end of the frequency converter to quickly and stably reach the set value, with better operation control effect on the output device, lower energy consumption, and effectively ensuring that the device can operate with high precision and high efficiency under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings. The drawings are schematic and should not be construed as imposing any limitation on the present invention. In the drawings:

[0059] Figure 1 FIG. shows the flowchart of a frequency converter control method based on a fuzzy neural network in Embodiment 1 of the present invention;

[0060] Figure 2 FIG. shows the architecture diagram of a sensor, a fuzzy neural network, a PID controller, a frequency converter, and an output device in Embodiment 1 of the present invention;

[0061] Figure 3 FIG. shows the architecture diagram of the fuzzy neural network in Embodiment 1 of the present invention;

[0062] Figure 4 FIG. shows the structural diagram of a frequency converter control system based on a fuzzy neural network in Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0064] Embodiment 1

[0065] This embodiment provides a frequency converter control method based on a fuzzy neural network, as Figure 1 shown, the method includes:

[0066] S1: Obtain sensor perception data in real time, and input the sensor perception data into the fuzzy neural network, and successively use the fuzzy neural network and the PID controller to obtain the output value of the frequency converter; wherein, the PID controller is arranged between the fuzzy neural network and the frequency converter;

[0067] S2: Back-propagate the output value of the frequency converter into the fuzzy neural network, and optimize the control parameter value of the PID controller according to the output value of the frequency converter to obtain the PID optimized parameter value;

[0068] S3: Use the PID controller to obtain a frequency converter adjustment signal according to the PID optimized parameter value;

[0069] S4: Control the output device of the frequency converter according to the frequency converter adjustment signal.

[0070] In this embodiment, first, the output value of the frequency converter is obtained through the fuzzy neural network and the PID controller to realize the forward propagation of the frequency converter control. Then, the obtained output value of the frequency converter is fed back and input into the fuzzy neural network. Using the backpropagation of the fuzzy neural network, parameter adjustment is performed based on the actually collected output value of the frequency converter and the set value and the error between the two, so as to optimize the control parameter value of the PID controller and obtain the PID optimized parameter value; the optimized PID optimized parameter value is then transmitted to the PID controller to realize the self-adjustment of the parameters of the PID controller, so that the PID controller has the ability of adaptive adjustment control; under the control of the adaptive adjustment PID controller, a frequency converter adjustment signal is obtained, and finally, the output device of the frequency converter is controlled by using this frequency converter adjustment signal, which can effectively realize the adaptive control of the frequency converter, so that the output device set at the output end of the frequency converter can quickly and stably reach the set value, and the operation control effect on the output device is better, with lower energy consumption, and it can effectively ensure that the device can operate with high precision and high efficiency under complex working conditions.

[0071] Next, each step of the frequency converter control method based on the fuzzy neural network in this embodiment will be further described.

[0072] In this embodiment, the output device of the frequency converter can be a motor configured for devices in the fields of industry, transportation, energy, household appliances, etc. In this embodiment, a motor configured for an oil fume machine in the household appliance field is taken as an example. This device is equipped with a sensor, specifically an oil fume sensor, which can perceive data such as the amount of oil fume in real time.

[0073] Specifically, in this embodiment, the architecture diagrams of the sensor, the fuzzy neural network, the PID controller, the frequency converter, and the output device are as Figure 2 shown.

[0074] In this embodiment S1, the sensor perception data is input into the fuzzy neural network, and the fuzzy neural network and the PID controller are used in sequence to obtain the output value of the frequency converter, including:

[0075] S11: Fuzzify the sensor perception data to obtain a fuzzy linguistic variable;

[0076] S12: Input the fuzzy linguistic variable into the fuzzy neural network to obtain fuzzy rules;

[0077] S13: Perform fuzzy inference operations according to the fuzzy rules to obtain the fuzzy adjustment amount corresponding to the PID controller;

[0078] S14: Defuzzify the fuzzy adjustment amount to obtain the control parameter value of the PID controller;

[0079] S15: Input the control parameter value of the PID controller into the PID controller to obtain a frequency converter control signal;

[0080] S16: Input the frequency converter control signal into the frequency converter, and collect the output value of the frequency converter.

[0081] The above is the forward propagation process for realizing the control of the frequency converter based on the fuzzy neural network and the PID controller. Fuzzifying the sensor perception data can convert precise numerical information into fuzzy information that is more in line with human thinking patterns. The obtained fuzzy linguistic variable helps to use fuzzy logic for reasoning and decision-making; inputting this fuzzy linguistic variable into the fuzzy neural network can automatically learn and extract the features and laws in the data, and optimize the processing and mapping of the input data by adjusting the weights, so as to obtain the fuzzy rules corresponding to the input; then perform fuzzy inference operations according to the obtained fuzzy rules, and the fuzzy adjustment amount for PID control can be accurately and efficiently deduced; through the defuzzification of this fuzzy adjustment amount, the fuzzy result obtained by fuzzy inference can be converted into an accurate numerical value, so that it can be directly used to adjust the parameters of the PID controller; inputting the obtained control parameter value into the PID controller can generate a frequency converter control signal to realize the PID control of the frequency converter; under the control of the generated frequency converter control signal, the frequency converter can output the corresponding frequency converter output value, and the configured sensor can be used to collect this frequency converter output value.

[0082] Through the forward propagation process based on the fuzzy neural network and the PID controller in this embodiment, an accurate and real-time output value of the frequency converter can be obtained, which is convenient for subsequent comparison with the set value using the actual output value to obtain an error, and using the error as a feedback signal to propagate backward to the fuzzy neural network, and using the backpropagation method to optimize and adjust the fuzzy rules of the fuzzy neural network and the control parameter values of the PID controller, so as to obtain an output result closer to the desired value, thereby realizing high-precision and high-efficiency control of the frequency converter.

[0083] Specifically, in step S11 of this embodiment, the sensor perception data is fuzzified to obtain fuzzy language variables, which can be fuzzy languages such as "large", "medium", and "small".

[0084] Preferably, the fuzzy neural network includes an input layer, an inference layer, and an output layer arranged in sequence;

[0085] Then in step S2 of this embodiment, the control parameter values of the PID controller are optimized according to the output value of the frequency converter to obtain PID optimized parameter values, including:

[0086] S21: Input the output value of the frequency converter into the input layer as the input variable of the input layer; and use the input layer to calculate a deviation function according to the input variable of the input layer;

[0087] S22: Input the deviation function into the inference layer, and calculate the input variable of the inference layer according to the deviation function; and use the inference layer to calculate the output function of the inference layer according to the input variable of the inference layer;

[0088] S23: Input the output function of the inference layer into the output layer, calculate the input variable of the output layer according to the output function of the inference layer; and use the output layer to calculate the PID optimized parameter values according to the input variable of the output layer.

[0089] The fuzzy neural network with the above architecture combines the advantages of fuzzy logic and neural networks, can be applied to data with fuzziness or uncertainty, and can handle complex non-linear mapping relationships; the backpropagation method of the fuzzy neural network based on this architecture can make the network output as close to the target value as possible, can automatically adjust the network parameters, and has high automation; through its layer-by-layer propagation, the entire network has significant advantages in dealing with high-dimensional data and complex pattern recognition problems, and has high flexibility and versatility.

[0090] It should be understood that the processes of sensor perception data, fuzzy neural network processing data, and using the PID controller to control the frequency converter are all real-time, and these real-time processes are all processed with the sampling period as the cycle.

[0091] Specifically, in the above step S21, the calculation formula of the deviation function is specifically:

[0092] e(t) = Er(t) - Eout(t);

[0093] where e(t) is the output function of the input layer at the t-th sampling period, representing the deviation function at the t-th sampling period; Eout(t) is the input variable of the input layer at the t-th sampling period, representing the output value of the frequency converter at the t-th sampling period; Er(t) is the set value corresponding to the output end of the frequency converter at the t-th sampling period.

[0094] The architecture diagram of the fuzzy neural network in this embodiment is as Figure 3 shown. In this network, there are specifically 2 input variables in the input layer, including the output value Eout(t) of the frequency converter and also Er(t); there are 5 input variables in the inference layer, and it will calculate 5 parameters as the input variables of the output layer; the output layer will calculate 3 parameters and output them to the PID controller.

[0095] Furthermore, the deviation function e(t) representing the t-th sampling period specifically refers to the difference between the actual sensed value of the oil fume volume and the set value of the oil fume volume at the t-th sampling period in this embodiment.

[0096] Specifically, in the above step S22, there are multiple first input parameters corresponding to the input variables of the inference layer at the t-th sampling period. Correspondingly, there are also multiple first output parameters corresponding to the output function of the inference layer at the t-th sampling period, and the number of the first output parameters is the same as the number of the first input parameters;

[0097] The specific formula for calculating the input variables of the inference layer according to the deviation function at the t-th sampling period is:

[0098]

[0099] where is the x-th first input parameter among the input variables of the inference layer at the t-th sampling period, n1 is the first iteration number corresponding to the input variables of the inference layer at the t-th sampling period, e k (t) is the value of the deviation function at the k-th iteration at the t-th sampling period, is the connection weight of the inference layer of the x-th first input parameter at the k-th iteration at the t-th sampling period;

[0100] The specific formula for calculating the output function of the inference layer according to the input variables of the inference layer at the t-th sampling period is:

[0101]

[0102] Among them, is the x-th first output parameter in the output function of the inference layer at the t-th sampling period, and tanh(·) is the hyperbolic tangent function.

[0103] Specifically, in the above step S23, there are multiple second input parameters corresponding to the input variables of the output layer at the t-th sampling period. Correspondingly, there are also multiple second output parameters corresponding to the output function of the output layer at the t-th sampling period, and the number of the second output parameters is the same as the number of the second input parameters;

[0104] The specific formula for calculating the input variables of the output layer according to the output function of the inference layer at the t-th sampling period is:

[0105]

[0106] Among them, is the y-th second input parameter in the input variables of the output layer at the t-th sampling period, n2 is the second iteration number corresponding to the input variables of the output layer at the t-th sampling period, is the y-th first output parameter in the output function of the inference layer at the t-th sampling period, is the output layer connection weight of the y-th second input parameter in the j-th iteration at the t-th sampling period;

[0107] The specific formula for calculating the output function of the output layer according to the input variables of the output layer at the t-th sampling period is:

[0108]

[0109] Among them, is the y-th second output parameter in the output function of the output layer at the t-th sampling period, representing the y-th optimization parameter value in the PID optimization parameter values at the t-th sampling period; tanh(·) is the hyperbolic tangent function.

[0110] Specifically, the PID optimization parameter values include an optimization proportionality coefficient, an optimization integral coefficient, and an optimization differential coefficient;

[0111] The three parameters output by the above output layer respectively correspond to the optimization proportionality coefficient, the optimization integral coefficient, and the optimization differential coefficient in the PID optimization parameter values;

[0112] That is, the optimization proportionality coefficient K p (t) is specifically f1 (3) (t), and the optimization integral coefficient K i (t) is specifically The optimized differential coefficient K at the t-th sampling period d (t) is specifically

[0113] Further, the objective function of the above fuzzy neural network can be set as a quadratic function of the deviation function, that is:

[0114]

[0115] where E(t) is the objective function of the fuzzy neural network.

[0116] Specifically, in this embodiment S3, the obtained optimized proportional coefficient, optimized integral coefficient, and optimized differential coefficient are input into the PID controller, and the inverter adjustment signal can be obtained.

[0117] where the inverter adjustment signal is specifically the equipment load corresponding to the output terminal device of the inverter;

[0118] Then, according to the PID optimization parameter values at the t-th sampling period, the specific formula for calculating the inverter adjustment signal at the t-th sampling period is:

[0119] U(t) = K p (t)·e(t) + K i (t)·∑e(t) + K d (t)·[e(t) - e(t - 1)];

[0120] where U(t) is the equipment load at the t-th sampling period, and e(t - 1) is the deviation function at the (t - 1)-th sampling period.

[0121] In the above formula, the deviation function at the (t - 1)-th sampling period specifically refers to the difference between the actual sensed value of the oil fume amount and the set value of the oil fume amount at the (t - 1)-th sampling period in this embodiment.

[0122] In this embodiment, the obtained optimized proportional coefficient, optimized integral coefficient, and optimized differential coefficient are input into the PID controller. According to the differences between the actual sensed values of the oil fume amount and the set value of the oil fume amount at the t-th sampling period and the (t - 1)-th sampling period, the equipment load U(t) at the t-th sampling period can be calculated; then, according to this equipment load U(t), the control operation of the inverter can be determined.

[0123] Preferably, the output terminal device of the inverter has an actual collected load;

[0124] This embodiment S4 includes:

[0125] Comparing the equipment load with the actual collected load;

[0126] When the device load is greater than the actual collected load, use the frequency converter to perform a loading operation on the output device; when the device load is less than the actual collected load, use the frequency converter to perform a load reduction operation on the output device; when the device load is equal to the actual collected load, maintain the current operation of the frequency converter.

[0127] By directly comparing the device load output by the frequency converter with the corresponding actual collected load, it is possible to efficiently determine whether it is necessary to use the frequency converter to load or unload the output device, which conforms to the actual usage conditions of the frequency converter and the output device, and can ensure the high-precision and high-efficiency operation of the frequency converter.

[0128] When the device load is greater than the actual collected load, it indicates that the current output device has not reached the ideal load state, and it is necessary to use the frequency converter controller to perform a loading operation (i.e., increase the load); when the device load is less than the actual collected load, it indicates that the current output device has exceeded the ideal load state, and it is necessary to use the frequency converter controller to perform a load reduction operation (i.e., reduce the load); when the device load is equal to the actual collected load, it indicates that the current output device is just in the ideal load state, and it is only necessary to maintain the current operation of the frequency converter.

[0129] Embodiment 2

[0130] A frequency converter control system based on a fuzzy neural network is applied to the frequency converter control method based on a fuzzy neural network in Embodiment 1, as Figure 4 shown, the system includes:

[0131] A forward propagation module, configured to obtain sensor perception data in real time, input the sensor perception data into the fuzzy neural network, and sequentially use the fuzzy neural network and the PID controller to obtain a frequency converter output value; wherein, the PID controller is arranged between the fuzzy neural network and the frequency converter;

[0132] A feedback adjustment module, configured to back-propagate the frequency converter output value into the fuzzy neural network, optimize the control parameter value of the PID controller according to the frequency converter output value, and obtain a PID optimized parameter value;

[0133] A PID optimization module, configured to use the PID controller to obtain a frequency converter adjustment signal according to the PID optimized parameter value;

[0134] A frequency converter control module, configured to control the output device of the frequency converter according to the frequency converter adjustment signal.

[0135] In this embodiment, first, the forward propagation module is used to obtain the output value of the frequency converter through the fuzzy neural network and the PID controller, realizing the forward propagation of the frequency converter control. Then, the feedback adjustment module is used to feedback the obtained output value of the frequency converter into the fuzzy neural network again. Using the backpropagation of the fuzzy neural network, parameter adjustment is performed based on the actually collected output value and set value of the frequency converter and the error between the two, realizing the optimization of the control parameter value of the PID controller and obtaining the optimized PID parameter value. Then, the PID optimization module is used to transfer the optimized PID parameter value to the PID controller again, realizing the self-adjustment of the parameters of the PID controller, so that the PID controller has the ability of adaptive adjustment control. Under the control of the adaptive adjustment PID controller, a frequency converter adjustment signal is obtained. Finally, through the frequency converter control module, the output device at the output end of the frequency converter is controlled by using this frequency converter adjustment signal, which can effectively realize the adaptive control of the frequency converter, enabling the output device at the output end of the frequency converter to quickly stabilize at the set value, with a more excellent operation control effect on the output device, lower energy consumption, and effectively ensuring that the device can also operate with high precision and high efficiency under complex working conditions.

[0136] In the frequency converter control system based on the fuzzy neural network described in this embodiment, the functions of each module are the same as the method steps of the frequency converter control method based on the fuzzy neural network described in Embodiment 1. Therefore, for the details not described in this embodiment, please refer to Embodiment 1 and Figure 1 and Figure 3 the specific description of, which will not be elaborated here.

[0137] Embodiment 3

[0138] This embodiment also provides a frequency converter control device based on a fuzzy neural network, including a processor, a memory, and a computer program stored in the memory and operable on the processor. When the computer program runs, it realizes the method steps in the frequency converter control method based on the fuzzy neural network in Embodiment 1.

[0139] By means of the computer program stored in the memory and running on the processor, the adaptive control of the frequency converter can be effectively realized, enabling the output device at the output end of the frequency converter to quickly stabilize at the set value, with a more excellent operation control effect on the output device, lower energy consumption, and effectively ensuring that the device can also operate with high precision and high efficiency under complex working conditions.

[0140] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the computer device, and connects various parts of the entire computer device through various interfaces and lines.

[0141] The memory can be used to store computer programs and / or models. The processor realizes various functions of the computer device by running or executing the computer programs and / or models stored in the memory, and by calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, video data, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a SmartMedia Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0142] It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by a computer program. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0143] These computer programs can also be stored in a computer-readable memory that can guide the computer or other programmable data processing devices to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device realizes the functions in the process Figure 1one process or multiple processes and / or boxes Figure 1 the functions specified in one box or multiple boxes.

[0144] These computer programs can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process or multiple processes and / or boxes Figure 1 one process or multiple processes and / or boxes Figure 1 the steps for implementing the functions specified in one box or multiple boxes.

[0145] This embodiment also provides a computer storage medium, which includes: at least one instruction that, when executed by a computer, implements the method steps in the inverter control method based on a fuzzy neural network in Embodiment 1.

[0146] By executing a computer storage medium containing at least one instruction, the adaptive control of the inverter can be effectively achieved, enabling the output device provided at the output end of the inverter to quickly and stably reach the set value. The operation control effect on the output device is more excellent, with lower energy consumption, and it can effectively ensure that the device operates with high precision and high efficiency under complex working conditions.

[0147] Similarly, for the details not described in this embodiment, refer to the specific descriptions in Embodiment 1, Embodiment 2 and Figures 1 to 3 which will not be elaborated here.

[0148] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A frequency converter control method based on a fuzzy neural network, characterized in that, The method includes: Obtaining sensor perception data in real time, and inputting the sensor perception data into a fuzzy neural network, and successively using the fuzzy neural network and a PID controller to obtain an output value of the frequency converter; wherein, the PID controller is arranged between the fuzzy neural network and the frequency converter; Backpropagating and inputting the output value of the frequency converter into the fuzzy neural network, and optimizing the control parameter value of the PID controller according to the output value of the frequency converter to obtain a PID optimization parameter value; Using the PID controller, according to the PID optimization parameter value, to obtain a frequency converter adjustment signal; Controlling the output-end device of the frequency converter according to the frequency converter adjustment signal.

2. The frequency converter control method based on a fuzzy neural network according to claim 1, characterized in that, The fuzzy neural network includes an input layer, an inference layer, and an output layer arranged in sequence; The optimizing the control parameter value of the PID controller according to the output value of the frequency converter to obtain a PID optimization parameter value includes: Inputting the output value of the frequency converter into the input layer as an input variable of the input layer; and using the input layer to calculate a deviation function according to the input variable of the input layer; Inputting the deviation function into the inference layer, and calculating an input variable of the inference layer according to the deviation function; and using the inference layer to calculate an output function of the inference layer according to the input variable of the inference layer; Inputting the output function of the inference layer into the output layer, and calculating an input variable of the output layer according to the output function of the inference layer; and using the output layer to calculate the PID optimization parameter value according to the input variable of the output layer.

3. The frequency converter control method based on a fuzzy neural network according to claim 2, characterized in that The specific calculation formula of the deviation function is: e(t) = Er(t) - Eout(t); wherein, e(t) is the output function of the input layer at the t-th sampling period, representing the deviation function at the t-th sampling period; Eout(t) is the input variable of the input layer at the t-th sampling period, representing the output value of the frequency converter at the t-th sampling period; Er(t) is the set value corresponding to the output end of the frequency converter at the t-th sampling period.

4. The frequency converter control method based on a fuzzy neural network according to claim 3, characterized in that There are multiple first input parameters corresponding to the input variable of the inference layer at the t-th sampling period. Correspondingly, there are also multiple first output parameters corresponding to the output function of the inference layer at the t-th sampling period, and the number of the first output parameters is the same as the number of the first input parameters; The specific formula for calculating the input variable of the inference layer according to the deviation function at the t-th sampling period is: Among them, is the x-th first input parameter among the input variables of the inference layer in the t-th sampling period, n1 is the first iteration number corresponding to the input variables of the inference layer in the t-th sampling period, and e k (t) is the value of the deviation function in the k-th iteration in the t-th sampling period, is the connection weight of the inference layer of the x-th first input parameter in the k-th iteration in the t-th sampling period; The specific formula for calculating the output function of the inference layer according to the input variable of the inference layer at the t-th sampling period is: Among them, is the x-th first output parameter in the output function of the inference layer at the t-th sampling period, and tanh(·) is the hyperbolic tangent function.

5. The frequency converter control method based on a fuzzy neural network according to claim 3, wherein There are multiple second input parameters corresponding to the input variable of the output layer at the t-th sampling period. Correspondingly, there are also multiple second output parameters corresponding to the output function of the output layer at the t-th sampling period, and the number of the second output parameters is the same as the number of the second input parameters; The specific formula for calculating the input variable of the output layer according to the output function of the inference layer at the t-th sampling period is: Among them, is the y-th second input parameter in the input variables of the output layer in the t-th sampling period, and n2 is the second iteration number corresponding to the input variables of the output layer in the t-th sampling period. is the y-th first output parameter in the output function of the inference layer in the t-th sampling period. is the connection weight of the output layer of the y-th second input parameter in the j-th iteration in the t-th sampling period. The specific formula for calculating the output function of the output layer according to the input variable of the output layer at the t-th sampling period is: Among them, is the y-th second output parameter in the output layer output function at the t-th sampling period, representing the y-th optimization parameter value among the PID optimization parameter values at the t-th sampling period; tanh(·) is the hyperbolic tangent function.

6. The frequency converter control method based on a fuzzy neural network according to claim 5, characterized in that, The PID optimization parameter values include an optimized proportional coefficient, an optimized integral coefficient, and an optimized derivative coefficient; Then the optimization proportionality coefficient K p at the t-th sampling period is specifically f1 (3) (t), and the optimization integral coefficient K i at the t-th sampling period is specifically The optimization differential coefficient K d at the t-th sampling period is specifically 7. The frequency converter control method based on a fuzzy neural network according to claim 6, characterized in that, The inverter adjustment signal is specifically the equipment load corresponding to the output terminal device of the inverter; Then, according to the PID optimization parameter values in the t-th sampling period, the specific formula for calculating the inverter adjustment signal in the t-th sampling period is: U(t) = K p (t)·e(t) + K i (t)·∑e(t) + K d (t)·[e(t) - e(t - 1)]; where U(t) is the equipment load in the t-th sampling period, and e(t - 1) is the deviation function in the (t - 1)-th sampling period.

8. The frequency converter control method based on a fuzzy neural network according to claim 7, characterized in that The output terminal device of the inverter has an actual collected load; Controlling the output terminal device of the inverter according to the inverter adjustment signal includes: Comparing the equipment load with the actual collected load; When the equipment load is greater than the actual collected load, using the inverter to perform a loading operation on the output terminal device; when the equipment load is less than the actual collected load, using the inverter to perform a unloading operation on the output terminal device; when the equipment load is equal to the actual collected load, maintaining the current operation of the inverter.

9. The frequency converter control method based on a fuzzy neural network according to any one of claims 1 to 8, characterized in that, Inputting the sensor perception data into a fuzzy neural network, and successively using the fuzzy neural network and a PID controller to obtain an inverter output value, including: Performing fuzzy processing on the sensor perception data to obtain a fuzzy linguistic variable; Inputting the fuzzy linguistic variable into the fuzzy neural network to obtain fuzzy rules; Performing fuzzy inference operation according to the fuzzy rules to obtain a fuzzy adjustment amount corresponding to the PID controller; Defuzzifying the fuzzy adjustment amount to obtain the control parameter value of the PID controller; Inputting the control parameter value of the PID controller into the PID controller to obtain an inverter control signal; Inputting the inverter control signal into the inverter to collect the inverter output value.

10. A frequency converter control system based on a fuzzy neural network, characterized in that, Applied to the inverter control method based on a fuzzy neural network according to any one of claims 1 to 9, the system includes: A forward propagation module, configured to obtain sensor perception data in real time, and input the sensor perception data into a fuzzy neural network, and successively use the fuzzy neural network and a PID controller to obtain an inverter output value; wherein, the PID controller is arranged between the fuzzy neural network and the inverter; A feedback adjustment module, configured to input the inverter output value in a reverse propagation manner into the fuzzy neural network, and optimize the control parameter value of the PID controller according to the inverter output value to obtain PID optimization parameter values; A PID optimization module, configured to use the PID controller to obtain an inverter adjustment signal according to the PID optimization parameter values; An inverter control module, configured to control the output terminal device of the inverter according to the inverter adjustment signal.

11. A frequency converter control device based on a fuzzy neural network, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program runs, it implements the method steps in the inverter control method based on a fuzzy neural network according to any one of claims 1 to 9.

12. A computer storage medium, characterized in that, The computer storage medium includes: at least one instruction, which, when executed by a computer, implements the method steps in the inverter control method based on a fuzzy neural network according to any one of claims 1 to 9.

Images