A method and system for constructing a non-intrusive load discrimination model

Abstract

The application provides a method and system for constructing a non-intrusive load identification model, which sequentially establishes single simulation models for different types of variable frequency motors, simulates operation in a first simulation system to obtain characteristic parameters of simulation waveforms of transformer outputs, calculates a fitting degree by comparing the characteristic parameters with actually recorded characteristic parameters of the transformer output waveforms, and determines an optimal simulation model of a single variable frequency motor; the optimal simulation models of different types of variable frequency motors are connected in parallel, and are operated in a second simulation system to obtain a large amount of simulation data, which is used as input of an initial non-intrusive load identification model for training, and when an identification accuracy reaches a set accuracy threshold, an optimal non-intrusive load identification model meeting the requirements is obtained. The method and system realize expansion of small data samples of actual operation of multiple variable frequency motors, and greatly improve the identification accuracy of the non-intrusive load identification model.

Description

Technical Field

[0001] This invention relates to the field of power load monitoring and signal processing, and more specifically, to a method and system for constructing a non-intrusive load identification model. Background Technology

[0002] Load identification technology provides fine-grained load energy consumption information for the power grid and users, playing a catalytic role in ensuring user power safety and improving power efficiency. Among the numerous user power loads, variable frequency motors account for the largest proportion. Compared with intervening load identification technology, non-intervening load identification technology has advantages such as low cost and ease of maintenance. Therefore, research on non-intervening load identification technology for variable frequency motors is currently a research hotspot. Most existing non-intervening load identification models require a large amount of actual operating data of power loads for training, and they are insufficient in terms of unknown load identification and generalization ability. For example, the operating characteristics of different types of variable frequency motors are highly similar, and it is even more difficult to distinguish them when different types of motors are mixed together. To address these issues, the traditional approach is to combine intervening and non-intervening load identification technologies. First, intervening equipment is used to record the waveform of individual devices and extract their operating characteristics to establish a load feature library. Then, non-intervening load identification technology is used to decompose and extract features from the bus waveform, and finally, it is matched with the load feature library. This method has two drawbacks: firstly, it is costly in terms of manpower and resources; secondly, the amount of data collected by the interventional equipment is limited, making it difficult to record waveforms for all types of variable frequency motors, resulting in low identification accuracy. Therefore, there is an urgent need for a method to expand the small sample data for variable frequency motors to improve the accuracy of non-intrusive load identification algorithms. Summary of the Invention

[0003] To address the high manpower and material costs and low identification accuracy caused by the combination of interventional and non-interventional load identification technologies in existing non-intrusive load identification model construction, which makes it difficult to record waveforms for all types of variable frequency motors, this invention provides a method and system for extracting and analyzing waveform feature parameters of on-load tap changers. This enables fully automatic, efficient, and high-precision extraction and analysis of waveform feature parameters of on-load tap changers, achieving comprehensive monitoring of the entire test waveform.

[0004] According to one aspect of the present invention, a method for constructing a non-intrusive load identification model is provided, the method comprising:

[0005] Establish an initial simulation model for a single variable frequency motor;

[0006] Based on the initial simulation model, a first simulation system for generator-transmission line-transformer-transmission line-variable frequency motor is established.

[0007] The operating state of the variable frequency motor is simulated in the first simulation system. The output waveform of the transformer in the first simulation system is extracted as the first bus-side simulation waveform, and the characteristic parameters of the first bus-side simulation waveform are calculated.

[0008] The feature parameters of the simulated waveform on the first bus side are compared with the feature parameters of the actual waveform on the first bus side that are actually recorded, and the fit is calculated.

[0009] The initial simulation model whose fitting degree is not less than the set matching degree threshold is determined as the optimal simulation model;

[0010] The optimal simulation models of multiple different types of variable frequency motors are connected in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor;

[0011] The second simulation system simulates the operation of multiple variable frequency motors running simultaneously. The output waveform of the transformer in the second simulation system is extracted as the simulation waveform on the second bus side, and the characteristic parameters of the simulation waveform on the second bus side are calculated.

[0012] The characteristic parameters of the second bus-side simulation waveform are used as input to the initial non-intrusive load identification algorithm model based on deep learning. The output of the initial non-intrusive load identification algorithm model is the number and type of variable frequency motors.

[0013] The identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system.

[0014] When the identification accuracy is not less than the set accuracy threshold, the initial non-intrusive load identification algorithm model is determined to be the optimal non-intrusive load identification algorithm model.

[0015] Optionally, the initial simulation model of a single variable frequency motor is established, wherein the initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

[0016] Optionally, the step of simulating the operating state of the variable frequency motor in the first simulation system, extracting the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculating the characteristic parameters of the first bus-side simulation waveform includes:

[0017] Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows:

[0018]

[0019] In the formula, ω is the angular frequency, j and k are the sampled value indices, and u k and u k-1 These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period;

[0020] The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows:

[0021]

[0022]

[0023] In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order;

[0024] Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows:

[0025] I p =max(i(k)),0 <k<K

[0026]

[0027] In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period.

[0028] The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows:

[0029]

[0030] In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

[0031] Optionally, the step of comparing the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that was actually recorded, and calculating the fitting degree r, is as follows:

[0032]

[0033]

[0034] In the formula, M represents the total number of characteristic parameters. y is the value of the i-th characteristic parameter of the actual waveform on the first bus side. i e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. i for With y i The residual.

[0035] Optionally, the identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows:

[0036]

[0037] In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

[0038] Optionally, the method further includes: when the fitting degree is less than a set matching degree threshold, adjusting the parameters of the initial simulation model and resimulating the operating state of the variable frequency motor in the first simulation system, wherein adjusting the parameters of the initial simulation model includes:

[0039] The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor.

[0040] In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor.

[0041] In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is:

[0042]

[0043] In the formula, G c (s) is the controller output, K p T is the proportionality coefficient. i and T d These are the time for integration and differentiation, respectively;

[0044] In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

[0045] Optionally, the method further includes:

[0046] When the identification accuracy is less than the set accuracy threshold, the parameters of the initial non-intrusive load identification algorithm model are adjusted, and the model after parameter adjustment is updated to the initial non-intrusive load identification algorithm model. The feature parameters of the second bus-side simulation waveform are then used as the input of the initial non-intrusive load identification algorithm model based on deep learning.

[0047] According to another aspect of the present invention, the present invention provides a system for constructing a non-intrusive load identification model, the system comprising:

[0048] The first model module is used to establish the initial simulation model of a single variable frequency motor;

[0049] The first simulation module is used to establish a first simulation system of generator-transmission line-transformer-transmission line-variable frequency motor based on the initial simulation model;

[0050] The first calculation module is used to simulate the operating state of the variable frequency motor in the first simulation system, extract the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculate the characteristic parameters of the first bus-side simulation waveform.

[0051] The second calculation module is used to compare the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that is actually recorded, and to calculate the fitting degree.

[0052] The second model module is used to determine the initial simulation model whose fitting degree is not less than the set matching degree threshold as the optimal simulation model;

[0053] The second simulation module is used to connect the optimal simulation models of multiple different types of variable frequency motors in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor;

[0054] The third calculation module is used to simulate the operating state of multiple variable frequency motors running simultaneously in the second simulation system, extract the output waveform of the transformer in the second simulation system as the second bus-side simulation waveform, and calculate the characteristic parameters of the second bus-side simulation waveform.

[0055] The model training module is used to take the feature parameters of the second bus-side simulation waveform as input to the initial non-intrusive load identification algorithm model based on deep learning, and obtain the output result of the initial non-intrusive load identification algorithm model, wherein the output result includes the number and type of variable frequency motors;

[0056] The fourth calculation module is used to determine the identification accuracy of the initial non-intrusive load identification algorithm model based on the output results and the number and types of the optimal simulation models of the variable frequency motor actually used in the second simulation system.

[0057] The optimal model module is used to determine the initial non-intrusive load identification algorithm model as the optimal non-intrusive load identification algorithm model when the identification accuracy is not less than the set accuracy threshold.

[0058] Optionally, the first model module establishes an initial simulation model of a single variable frequency motor, wherein the initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

[0059] Optionally, the first calculation module simulates the operating state of the variable frequency motor in the first simulation system, extracts the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculates the characteristic parameters of the first bus-side simulation waveform, including:

[0060] Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows:

[0061]

[0062] In the formula, ω is the angular frequency, j and k are the sampling value indices, and uk and uk -1 These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period;

[0063] The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows:

[0064]

[0065]

[0066] In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order;

[0067] Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows:

[0068] i p =max(i(k)),0 <k<K

[0069]

[0070] In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period.

[0071] The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows:

[0072]

[0073] In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

[0074] Optionally, the step of comparing the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that was actually recorded, and calculating the fitting degree r, is as follows:

[0075]

[0076]

[0077] In the formula, M represents the total number of characteristic parameters. y is the value of the i-th characteristic parameter of the actual waveform on the first bus side. i e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. ifor With y i The residual.

[0078] Optionally, the identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows:

[0079]

[0080] In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

[0081] Optionally, the method further includes: when the fitting degree is less than a set matching degree threshold, adjusting the parameters of the initial simulation model and returning to the first calculation module, wherein adjusting the parameters of the initial simulation model includes:

[0082] The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor.

[0083] In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor.

[0084] In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is:

[0085]

[0086] In the formula, G c (s) is the controller output, K p T is the proportionality coefficient. i and T d These are the time for integration and differentiation, respectively;

[0087] In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

[0088] Optionally, the system further includes a second adjustment module, used to adjust the parameters of the initial non-intrusive load identification algorithm model when the identification accuracy is less than a set accuracy threshold, and then update the model with adjusted parameters to the initial non-intrusive load identification algorithm model before returning to the model training module.

[0089] According to another aspect of the present invention, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-described methods for constructing a non-intrusive load identification model.

[0090] According to another aspect of the present invention, the present invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein:

[0091] The processor, communication interface, and memory communicate with each other through a communication bus;

[0092] Memory, used to store computer programs;

[0093] When a processor executes a program stored in memory, it implements the steps of any of the above methods for constructing a surface non-intrusive load identification model.

[0094] The method and system for constructing a non-intrusive load identification model according to the present invention include the following steps: First, establish individual simulation models for different types of variable frequency motors sequentially. Simulate the operation in a first simulation system to obtain characteristic parameters of the transformer output waveform. Calculate the fitting degree between these simulation parameters and the actual recorded transformer output waveform to determine the optimal simulation model for a single variable frequency motor. Second, connect the optimal simulation models of different types of variable frequency motors in parallel and run them in a second simulation system to obtain a large amount of simulation data. Use this simulation data as input to train an initial non-intrusive load identification model. When the identification accuracy reaches a set accuracy threshold, a satisfactory optimal non-intrusive load identification model is obtained. The method and system, by establishing an initial simulation model for a single variable frequency motor and simulating the actual operating state of the variable frequency motor by matching it with the actual operating waveform, expands the small data sample of multiple variable frequency motors' actual operation. The expanded sample is then used as input to train the non-intrusive load identification model, thereby significantly improving the identification accuracy of the non-intrusive load identification model of the variable frequency motor obtained through large-sample data training. Attached Figure Description

[0095] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures:

[0096] Figure 1A flowchart of a method for constructing a non-intrusive load identification model according to a preferred embodiment of the present invention;

[0097] Figure 2 This is a schematic diagram of the structure of a system for constructing a non-intrusive load identification model according to a preferred embodiment of the present invention. Detailed Implementation

[0098] Exemplary embodiments of the invention will now be described with reference to the accompanying drawings. However, the invention may be embodied in many different forms and is not limited to the embodiments described herein. These embodiments are provided to fully and completely disclose the invention and to fully convey its scope to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the drawings is not intended to limit the invention. In the drawings, the same units / elements are referred to by the same reference numerals.

[0099] Unless otherwise stated, the terms used herein (including technical terms) have their common meaning as understood by one of ordinary skill in the art. Furthermore, it is understood that terms defined in commonly used dictionaries should be understood to have a meaning consistent with the context of their relevant field, and not to be interpreted as having an idealized or overly formal meaning.

[0100] Example Method

[0101] Figure 1 This is a flowchart illustrating a method for constructing a non-intrusive load identification model according to a preferred embodiment of the present invention. Figure 1 As shown, the method for constructing a non-intrusive load identification model according to this preferred embodiment begins with step 101.

[0102] In step 101, an initial simulation model of a single variable frequency motor is established.

[0103] Preferably, the initial simulation model of a single variable frequency motor is established, wherein the initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

[0104] In this preferred embodiment, a simulation model of a single variable frequency motor is built based on the variable frequency principle of the variable frequency motor. The main body adopts a physical model, the parameter adjustment part adopts a mathematical model, and PID is used for automatic adjustment.

[0105] In step 102, based on the initial simulation model, a first simulation system of generator-transmission line-transformer-transmission line-variable frequency motor is established.

[0106] In step 103, the operating state of the variable frequency motor is simulated in the first simulation system, the output waveform of the transformer in the first simulation system is extracted as the first bus-side simulation waveform, and the characteristic parameters of the first bus-side simulation waveform are calculated.

[0107] Preferably, the step of simulating the operating state of the variable frequency motor in the first simulation system, extracting the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculating the characteristic parameters of the first bus-side simulation waveform includes:

[0108] Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows:

[0109]

[0110] In the formula, ω is the angular frequency, j and k are the sampled value indices, and u k and u k-1 These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period;

[0111] The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows:

[0112]

[0113]

[0114] In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order;

[0115] Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows:

[0116] I p =max(i(k)),0 <k<K

[0117]

[0118] In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period.

[0119] The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows:

[0120]

[0121] In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

[0122] In this preferred embodiment, the characteristic parameters of the actual waveform on the first bus side that is actually recorded are also calculated in the manner described above, and will not be repeated here.

[0123] In step 104, the feature parameters of the simulated waveform on the first bus side are compared with the feature parameters of the actual waveform on the first bus side that are actually recorded, and the fitting degree is calculated.

[0124] Preferably, the step of comparing the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side, and calculating the fitting degree r, is as follows:

[0125]

[0126] In the formula, M represents the total number of characteristic parameters. y is the value of the i-th characteristic parameter of the actual waveform on the first bus side. i e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. i for With y i The residual.

[0127] Preferably, the method further includes: when the fitting degree is less than a set matching degree threshold, adjusting the parameters of the initial simulation model and resimulating the operating state of the variable frequency motor in the first simulation system, wherein adjusting the parameters of the initial simulation model includes:

[0128] The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor.

[0129] In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor.

[0130] In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is:

[0131]

[0132] In the formula, Gc(s) is the controller output, Kp is the proportional coefficient, and Ti and Td are the integral and derivative times, respectively.

[0133] In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

[0134] In this preferred embodiment, the closer the fitting degree is to 1, the better the fitting performance of the established simulation model of a single variable frequency motor. When the fitting degree is less than the set matching degree threshold, it indicates that the fitting performance of the established simulation model cannot meet the requirements, therefore, the parameters of the simulation model need to be adjusted. This invention mainly involves adjusting control parameters such as excitation current, fundamental frequency, speed, and carrier amplitude in the simulation model to adjust output parameters such as reactive power, output power, and stator current of the variable frequency motor. This is because the control parameters and output parameters are independent of each other, making adjustment more convenient and faster. It should be noted that the adjustment of the stator current, an output parameter of the variable frequency motor, is achieved by automatically adjusting the speed of the variable frequency motor using PID control, and then changing the electromagnetic torque of the variable frequency motor by changing the amplitude of the carrier wave. The principle is as follows:

[0135]

[0136] In the formula, P is the motor power, F is the torque, V is the linear velocity, and T is the torque. e R is the electromagnetic torque, n is the radius of action, U is the motor speed, and I is the motor operating voltage.

[0137] As shown in the formula above, when the speed of the variable frequency motor is fixed, the stator current is only related to the electromagnetic torque. Therefore, if it is necessary to adjust the stator current, it is necessary to use PID control to make the motor's output speed match the target speed, and then change the electromagnetic torque of the variable frequency motor by changing the amplitude of the carrier wave, thus achieving the purpose of adjusting the stator current.

[0138] In step 105, the initial simulation model whose fitting degree is not less than the set matching degree threshold is determined as the optimal simulation model.

[0139] Since there are many types of variable frequency motors, in order to improve the identification accuracy of the non-intrusive load model, it is necessary to train a sufficient number of samples. Therefore, in this preferred embodiment, the optimal simulation model is actually established for each type of variable frequency motor as much as possible, in order to simulate more load operating states in the actual power grid.

[0140] Step 106: Connect the optimal simulation models of multiple different types of variable frequency motors in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor.

[0141] In step 107, the operating state of multiple variable frequency motors running simultaneously is simulated in the second simulation system. The output waveform of the transformer in the second simulation system is extracted as the second bus-side simulation waveform, and the characteristic parameters of the second bus-side simulation waveform are calculated.

[0142] In this preferred embodiment, the method for calculating the characteristic parameters of the second bus-side simulation waveform is the same as the method for calculating the characteristic parameters of the first bus-side simulation waveform, and will not be described again here.

[0143] In step 108, the feature parameters of the second bus-side simulation waveform are used as input to the initial non-intrusive load identification algorithm model based on deep learning. The output of the initial non-intrusive load identification algorithm model is the number and type of variable frequency motors.

[0144] In step 109, the identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of the optimal simulation models of the variable frequency motor actually used in the second simulation system.

[0145] Preferably, the identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows:

[0146]

[0147] In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

[0148] In this preferred embodiment, the number and type of variable frequency motors in the output results are statistically analyzed in different cases. For example, if a type of variable frequency motor exists in the output results but does not exist in the second simulation system, the number of variable frequency motors of that type is determined to be 0, regardless of the number of variable frequency motors of that type in the output results. If a type of variable frequency motor exists in both the output results and the second simulation system, the number of that type in the output results is not greater than the number in the second simulation system, and the number in the output results is used as the basis for statistical analysis. If the number of that type in the output results is greater than the number in the second simulation system, the number in the second simulation system is used as the basis for statistical analysis. Finally, the number of variable frequency motors of all types in the output results is summed to obtain the total number of variable frequency motors determined based on the output results.

[0149] In step 110, when the identification accuracy is not less than the set accuracy threshold, the initial non-intrusive load identification algorithm model is determined to be the optimal non-intrusive load identification algorithm model.

[0150] Preferably, the method further includes:

[0151] When the identification accuracy is less than the set accuracy threshold, adjust the parameters of the initial non-intrusive load identification algorithm model, update the adjusted model to the initial non-intrusive load identification algorithm model, and return to step 108.

[0152] In this preferred embodiment, when the identification accuracy determined by the output result of the initial non-intrusive load identification model is not up to the set accuracy threshold during training, it is necessary to adjust the parameters of the initial non-intrusive load identification model, use the adjusted model as the initial non-intrusive load identification model, and train again using simulation data obtained from the second simulation system as samples, thereby obtaining the optimal non-intrusive load identification model with higher identification accuracy.

[0153] Of course, in addition to adjusting the parameters of the initial non-intrusive load identification model to obtain the optimal non-intrusive load identification model, it is also possible to establish simulation models of more types of variable frequency motors to obtain more simulation data as input to the initial non-intrusive load identification model, and at the same time continuously optimize the initial non-intrusive load identification model to obtain the optimal non-intrusive load identification model.

[0154] The method and system for constructing a non-intrusive load identification model according to this preferred embodiment establishes individual simulation models for different types of variable frequency motors sequentially. The model is then run in a first simulation system to obtain the characteristic parameters of the transformer output waveform. The optimal simulation model for a single variable frequency motor is determined by calculating the fitting degree between the simulation model and the characteristic parameters of the actual transformer output waveform. The optimal simulation models for different types of variable frequency motors are then connected in parallel and run in a second simulation system to obtain a large amount of simulation data. This simulation data is used as input to train the initial non-intrusive load identification model. When the identification accuracy reaches a set accuracy threshold, the optimal non-intrusive load identification model that meets the requirements is obtained. The method and system establish an initial simulation model for a single variable frequency motor and simulate the actual operating state of the variable frequency motor by matching it with the actual operating waveform. This expands the small data sample of multiple variable frequency motors into a larger dataset. The expanded sample is then used as input to train the non-intrusive load identification model, thereby greatly improving the identification accuracy of the non-intrusive load identification model for variable frequency motors trained with large sample data.

[0155] Example System

[0156] Figure 2 This is a schematic diagram of the structure of a system for constructing a non-intrusive load identification model according to a preferred embodiment of the present invention. Figure 2 As shown, the system 200 for constructing a non-intrusive load identification model according to this preferred embodiment includes:

[0157] The first model module 201 is used to establish the initial simulation model of a single variable frequency motor;

[0158] The first simulation module 202 is used to establish a first simulation system of generator-transmission line-transformer-transmission line-variable frequency motor based on the initial simulation model;

[0159] The first calculation module 203 is used to simulate the operating state of the variable frequency motor in the first simulation system, extract the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculate the characteristic parameters of the first bus-side simulation waveform.

[0160] The second calculation module 204 is used to compare the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that is actually recorded, and to calculate the fitting degree.

[0161] The second model module 205 is used to determine the initial simulation model whose fitting degree is not less than the set matching degree threshold as the optimal simulation model;

[0162] The second simulation module 206 is used to connect the optimal simulation models of multiple different types of variable frequency motors in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor;

[0163] The third calculation module 207 is used to simulate the operating state of multiple variable frequency motors running simultaneously in the second simulation system, extract the output waveform of the transformer in the second simulation system as the second bus-side simulation waveform, and calculate the characteristic parameters of the second bus-side simulation waveform.

[0164] The model training module 208 is used to take the feature parameters of the second bus-side simulation waveform as input to the initial non-intrusive load identification algorithm model based on deep learning. The output of the initial non-intrusive load identification algorithm model is the number and type of variable frequency motors.

[0165] The fourth calculation module 209 is used to determine the identification accuracy of the initial non-intrusive load identification algorithm model based on the output results and the number and types of the optimal simulation models of the variable frequency motor actually used in the second simulation system.

[0166] The optimal model module 210 is used to determine the initial non-intrusive load identification algorithm model as the optimal non-intrusive load identification algorithm model when the identification accuracy is not less than the set accuracy threshold.

[0167] Preferably, the first model module 201 establishes an initial simulation model of a single variable frequency motor, wherein the initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

[0168] Preferably, the first calculation module 203 simulates the operating state of the variable frequency motor in the first simulation system, extracts the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculates the characteristic parameters of the first bus-side simulation waveform, including:

[0169] Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows:

[0170]

[0171] In the formula, ω is the angular frequency, j and k are the sampled value indices, and u k and u k-1These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period;

[0172] The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows:

[0173]

[0174]

[0175] In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order;

[0176] Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows:

[0177] l p =max(i(k)),0 <k<K

[0178]

[0179] In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period.

[0180] The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows:

[0181]

[0182] In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

[0183] Preferably, the second calculation module 204 compares the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that was actually recorded, and calculates the fitting degree r, the calculation formula of which is:

[0184]

[0185]

[0186] In the formula, M represents the total number of characteristic parameters. y is the value of the i-th characteristic parameter of the actual waveform on the first bus side. i e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. i for With y i The residual.

[0187] Preferably, the fourth calculation module 209 determines the identification accuracy of the initial non-intrusive load identification algorithm model based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows:

[0188]

[0189] In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

[0190] Preferably, the system further includes a first adjustment module, used to adjust the parameters of the initial simulation model when the fitting degree is less than a set matching degree threshold, and input the adjusted parameters into a second simulation module, wherein adjusting the parameters of the initial simulation model includes:

[0191] The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor.

[0192] In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor.

[0193] In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is:

[0194]

[0195] In the formula, G c (s) is the controller output, K p T is the proportionality coefficient. i and T d These are the time for integration and differentiation, respectively;

[0196] In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

[0197] Preferably, the system further includes a second adjustment module, used to adjust the parameters of the initial non-intrusive load identification algorithm model when the identification accuracy is less than a set accuracy threshold, and update the adjusted model to the initial non-intrusive load identification algorithm model.

[0198] The steps of the system for constructing a non-intrusive load identification model described in this preferred embodiment to generate simulated partial discharge signals based on real on-site partial discharge signals are the same as the steps in the method for constructing a non-intrusive load identification model described above, and the technical effects achieved are also the same, so they will not be repeated here.

[0199] A preferred embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in any of the above-described methods for constructing a non-intrusive load identification model.

[0200] A preferred embodiment of the present invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein:

[0201] The processor, communication interface, and memory communicate with each other through a communication bus;

[0202] Memory, used to store computer programs;

[0203] When the processor executes a program stored in memory, it implements the steps in any of the above methods for constructing a non-intrusive load identification model.

[0204] The invention has been described with reference to a few embodiments. However, as will be known to those skilled in the art, and as defined in the appended claims, other embodiments besides those disclosed above fall equivalently within the scope of the invention.

[0205] Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the art, unless otherwise expressly defined herein. All references to “a / the / the [device, component, etc.]” ​​are openly interpreted as at least one instance of said device, component, etc., unless otherwise expressly stated. The steps of any method disclosed herein need not be performed in the exact order disclosed unless explicitly stated otherwise.

[0206] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0207] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0208] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0209] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0210] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for a non-intrusive load identification model, characterized in that, include: Establish an initial simulation model for a single variable frequency motor; Based on the initial simulation model, a first simulation system for generator-transmission line-transformer-transmission line-variable frequency motor is established. The operating state of the variable frequency motor is simulated in the first simulation system. The output waveform of the transformer in the first simulation system is extracted as the first bus-side simulation waveform, and the characteristic parameters of the first bus-side simulation waveform are calculated. The feature parameters of the simulated waveform on the first bus side are compared with the feature parameters of the actual waveform on the first bus side that are actually recorded, and the fit is calculated. The initial simulation model whose fitting degree is not less than the set matching degree threshold is determined as the optimal simulation model; The optimal simulation models of multiple different types of variable frequency motors are connected in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor; The second simulation system simulates the operation of multiple variable frequency motors running simultaneously. The output waveform of the transformer in the second simulation system is extracted as the simulation waveform on the second bus side, and the characteristic parameters of the simulation waveform on the second bus side are calculated. The characteristic parameters of the second bus-side simulation waveform are used as input to the initial non-intrusive load identification algorithm model based on deep learning, and the output results of the initial non-intrusive load identification algorithm model are obtained, wherein the output results include the number and type of variable frequency motors; The identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. When the identification accuracy is not less than the set accuracy threshold, the initial non-intrusive load identification algorithm model is determined to be the optimal non-intrusive load identification algorithm model.

2. The method according to claim 1, characterized in that, The initial simulation model of a single variable frequency motor is established, wherein the initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

3. The method according to claim 1, characterized in that, The process of simulating the operating state of a variable frequency motor in the first simulation system, extracting the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculating the characteristic parameters of the first bus-side simulation waveform includes: Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, ω is the angular frequency, j and k are the sampled value indices, and u k and u k-1 These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period; The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order; Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows: In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period. The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

4. The method according to claim 1, characterized in that, The feature parameters of the simulated waveform on the first bus side are compared with the feature parameters of the actual waveform on the first bus side, and the fitting degree r is calculated. The calculation formula is as follows: In the formula, M represents the total number of characteristic parameters. The value of the i-th characteristic parameter of the actual waveform on the first bus side. e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. i for and The residual.

5. The method according to claim 1, characterized in that, The identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows: In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

6. The method according to claim 2, characterized in that, The method further includes: when the fitting degree is less than a set matching degree threshold, adjusting the parameters of the initial simulation model and resimulating the operating state of the variable frequency motor in the first simulation system, wherein adjusting the parameters of the initial simulation model includes: The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor. In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor. In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is: In the formula, G c (s) is the controller output, K p T is the proportionality coefficient. i and T d These are the time for integration and differentiation, respectively; In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

7. The method according to claim 1, characterized in that, The method further includes: When the identification accuracy is less than the set accuracy threshold, the parameters of the initial non-intrusive load identification algorithm model are adjusted, and the model after parameter adjustment is updated to the initial non-intrusive load identification algorithm model. The feature parameters of the second bus-side simulation waveform are then used as the input of the initial non-intrusive load identification algorithm model based on deep learning.

8. A system for constructing a non-intrusive load identification model, characterized in that, The system includes: The first model module is used to establish the initial simulation model of a single variable frequency motor; The first simulation module is used to establish a first simulation system of generator-transmission line-transformer-transmission line-variable frequency motor based on the initial simulation model; The first calculation module is used to simulate the operating state of the variable frequency motor in the first simulation system, extract the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculate the characteristic parameters of the first bus-side simulation waveform. The second calculation module is used to compare the feature parameters of the simulated waveform on the first bus side with the feature parameters of the actual waveform on the first bus side that is actually recorded, and to calculate the fitting degree. The second model module is used to determine the initial simulation model whose fitting degree is not less than the set matching degree threshold as the optimal simulation model; The second simulation module is used to connect the optimal simulation models of multiple different types of variable frequency motors in parallel to establish a second simulation system of generator-transmission line-transformer-transmission line-variable frequency motor; The third calculation module is used to simulate the operating state of multiple variable frequency motors running simultaneously in the second simulation system, extract the output waveform of the transformer in the second simulation system as the second bus-side simulation waveform, and calculate the characteristic parameters of the second bus-side simulation waveform. The model training module is used to take the feature parameters of the second bus-side simulation waveform as input to the initial non-intrusive load identification algorithm model based on deep learning, and obtain the output result of the initial non-intrusive load identification algorithm model, wherein the output result includes the number and type of variable frequency motors; The fourth calculation module is used to determine the identification accuracy of the initial non-intrusive load identification algorithm model based on the output results and the number and types of the optimal simulation models of the variable frequency motor actually used in the second simulation system. The optimal model module is used to determine the initial non-intrusive load identification algorithm model as the optimal non-intrusive load identification algorithm model when the identification accuracy is not less than the set accuracy threshold.

9. The system according to claim 8, characterized in that, The first model module establishes an initial simulation model of a single variable frequency motor. The initial simulation model includes an AC power input module, a rectifier module, an inverter module, a variable frequency motor operation module, and a parameter adjustment module. The inverter module consists of 6 thyristors, and an SPWM wave is formed by a sine wave fundamental and a triangular wave carrier to trigger the thyristors to conduct. The parameter adjustment module is used to adjust the speed of the variable frequency motor.

10. The system according to claim 8, characterized in that, The first calculation module simulates the operating state of the variable frequency motor in the first simulation system, extracts the output waveform of the transformer in the first simulation system as the first bus-side simulation waveform, and calculates the characteristic parameters of the first bus-side simulation waveform, including: Calculate the power factor angle of the first bus-side simulation waveform based on the voltage and current sample values ​​in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, ω is the angular frequency, j and k are the sampled value indices, and u k and u k-1 These are the voltage sampling values ​​for the k-th and (k-1)-th times, respectively, i k and i k-1 T represents the current sampling values ​​for the k-th and (k-1)-th times, respectively. s The sampling period; The active power P and reactive power Q of the first bus-side simulation waveform are calculated based on the voltage and current waveforms in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, P and Q represent active power and reactive power, respectively, and u n and i n These represent voltage and current, respectively; θ is the phase difference between voltage and current, and n is the harmonic order; Calculate the peak current I of the first bus-side simulation waveform based on the current waveform in the first bus-side simulation waveform. p and the root mean square value of current I rms The calculation formula is as follows: In the formula, K is the total number of samplings in all sampling periods, i(k) is the kth current sample value, N is the total number of samplings in a single sampling period, and i(n) is the nth current sample value in a single sampling period. The total harmonic content H of the current in the first bus-side simulation waveform is calculated based on the current waveform in the first bus-side simulation waveform. The calculation formula is as follows: In the formula, I i It is the amplitude of the i-th harmonic of the current in the frequency domain, and h is the order of the highest harmonic.

11. The system according to claim 8, characterized in that, The feature parameters of the simulated waveform on the first bus side are compared with the feature parameters of the actual waveform on the first bus side, and the fitting degree r is calculated. The calculation formula is as follows: In the formula, M represents the total number of characteristic parameters. The value of the i-th characteristic parameter of the actual waveform on the first bus side. e represents the value of the i-th characteristic parameter of the simulated waveform on the first bus side. i for and The residual.

12. The system according to claim 8, characterized in that, The identification accuracy of the initial non-intrusive load identification algorithm model is determined based on the output results and the number and types of optimal simulation models of the variable frequency motor actually used in the second simulation system. The calculation formula is as follows: In the formula, δ is the identification accuracy, O is the total number of variable frequency motors determined based on the output results, and N is the total number of optimal simulation models of variable frequency motors actually used in the second simulation system.

13. The system according to claim 9, characterized in that, The system further includes: when the fitting degree is less than a set matching degree threshold, adjusting the parameters of the initial simulation model and returning to the first calculation module, wherein adjusting the parameters of the initial simulation model includes: The excitation current of the variable frequency motor is adjusted in the variable frequency motor operation module to change the reactive power of the variable frequency motor. In the inverter module, a trigger signal is used to activate the thyristor to adjust the fundamental frequency and thus change the output power of the variable frequency motor. In the parameter adjustment module, the speed of the variable frequency motor is automatically adjusted based on an incremental algorithm using PID control. The PID control employs a PID controller determined by the Ziegler-Nichols first method. For a set fixed target speed S0 of the variable frequency motor, the difference between the intended output speed S and S0 is used as the input to the PID controller. The control parameters Kp, Ti, and Td of the PID controller are adjusted until S and S0 are equal. The expression of the PID controller is: In the formula, G c (s) is the controller output, K p T is the proportionality coefficient. i and T d These are the time for integration and differentiation, respectively; In the inverter module, a trigger signal that activates the thyristor is used to adjust the carrier amplitude to change the stator current of the variable frequency motor.

14. The system according to claim 8, characterized in that, The system also includes a second adjustment module, which is used to adjust the parameters of the initial non-intrusive load identification algorithm model when the identification accuracy is less than the set accuracy threshold, and then update the model with the adjusted parameters to the initial non-intrusive load identification algorithm model before returning to the model training module.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method described in any one of claims 1-7.

16. An electronic device, characterized in that, Includes a processor, communication interface, memory, and communication bus, among which: The processor, communication interface, and memory communicate with each other through a communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the method described in any one of claims 1-7.

Images