Equivalent load model parameter identification method, device, equipment, medium and product

Abstract

The invention provides an equivalent load model parameter identification method and device, equipment, a medium and a product, and the method comprises the steps: constructing an equivalent load model considering frequency characteristics, and selecting noise-like identification data; the method comprises the following steps: identifying noise-like identification data of an equivalent load model, classifying each to-be-identified parameter according to the identification sensitivity, and carrying out pre-identification-parameter regularization identification-variable torque load inertia identification step-by-step identification on the equivalent load model parameter according to the noise-like identification data and the identifiability of each to-be-identified parameter to obtain an identification result of the equivalent load model parameter. The equivalent load model structure considering the frequency characteristics can effectively describe the voltage / frequency comprehensive dynamic characteristics of the electric equipment; through pre-identification, parameter regularization identification and variable torque load inertia identification step-by-step decoupling identification, stable parameter identification under multi-input data identification is realized, various mechanical characteristics of a real motor load are adapted, a stable and reliable identification result is obtained, and identifiability of multiple parameters of an equivalent load model under noise-like identification data is improved.

Description

Technical Field

[0001] This invention relates to the field of power system load modeling technology, and in particular to a method, apparatus, equipment, medium and product for identifying equivalent load model parameters. Background Technology

[0002] Current equivalent load models in power system analysis primarily reflect the relationship between load output active / reactive power and input voltage. However, with the increasing proportion of frequency-sensitive devices such as air conditioners and energy storage in telecommunications power consumption scenarios, equivalent load models used for system analysis typically need to further encompass frequency-related parameters.

[0003] Therefore, considering the frequency characteristics of the equivalent load model, the increased input quantity and parameters to be identified will further increase the difficulty of identification. However, the existing identification methods for multi-parameter load models select key parameters for identification, which fails to fundamentally solve the problem of parameter identifiability.

[0004] Considering the diversification of information and communication equipment during the intelligent transformation of large data centers and industrial parks, the loads include not only those exhibiting pure power / impedance characteristics but also a large number of motor-driven loads such as air conditioners and water pumps. Furthermore, as information and communication loads participate as flexible resources in grid demand response and ancillary services (such as frequency response and frequency regulation ancillary services), the need for frequency response characteristic analysis of these devices increases. Therefore, from the perspectives of operational stability of information and communication equipment, power supply reliability, and information energy security, it is of great significance to construct a complete equivalent load model and identify model parameters based on measurement data. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, medium, and product for identifying equivalent load model parameters, which addresses the shortcomings of incomplete equivalent load modeling and the difficulty of multi-parameter identification in existing technologies. It enables the construction of a complete equivalent load model and the step-by-step decoupled identification of equivalent load model parameters, including pre-identification, parameter regularization identification, and identification of load inertia with varying torque, thereby improving the identifiability of multi-parameter identification under noise-like identification data.

[0006] This invention provides a method for identifying parameters of an equivalent load model, comprising: Construct an equivalent load model that considers frequency characteristics; Based on the structure of the equivalent load model, noise identification data is selected. The identifiability of each parameter to be identified is determined by classifying each parameter based on its identification sensitivity. Based on the noise identification data and the identifiability of each parameter to be identified, the equivalent load model parameters are identified in a step-by-step manner: pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification results of the equivalent load model parameters.

[0007] According to a method for identifying equivalent load model parameters provided by the present invention, the parameters to be identified include motor inertia parameters and electromagnetic parameters with higher identifiability than the motor inertia parameters. The electromagnetic parameters include frequency characteristic parameters and equivalent load electromagnetic parameters with higher identifiability than the frequency characteristic parameters. The method involves a step-by-step identification process based on the noise identification data and the identifiability of each parameter to be identified, performing pre-identification, parameter regularization identification, and variable torque load inertia identification on the equivalent load model parameters to obtain the identification results of the equivalent load model parameters. This process includes: Pre-identification is performed based on noise identification data to obtain the pre-identification results of electromagnetic parameters of equivalent load. Based on noise identification data, the pre-identification results of electromagnetic parameters of the equivalent load, and the frequency characteristic parameter identification with parameter regularization of the system frequency, the effective identification results of electromagnetic parameters are obtained. Based on the noise identification data, the effective identification results of the electromagnetic parameters, and the electromechanical state of the motor, the load inertia of the variable torque is identified, and the motor inertia parameter identification results are obtained. Based on the effective identification results of the electromagnetic parameters and the identification results of the motor inertia parameters, the identification results of the equivalent load model parameters are obtained.

[0008] According to the method for identifying equivalent load model parameters provided by the present invention, the noise-like identification data includes measured voltage input and measured power output; the pre-identification based on the noise-like identification data to obtain the pre-identification result of the equivalent load electromagnetic parameters includes: Based on the measured voltage input and measured power output at the current moment, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state. Based on the first predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the first predicted power. Based on the measured power and the first predicted power at the same moment, a first objective function is determined; the objective of the first objective function is to minimize the deviation between the measured power and the first predicted power. The first objective function is optimized using a differential evolution algorithm to obtain the pre-identification results of the electromagnetic parameters of the equivalent load.

[0009] According to the method for identifying equivalent load model parameters provided by the present invention, the step of predicting the induced electromotive force state at the next moment based on the measured voltage input and measured power output at the current moment to obtain a first predicted induced electromotive force state includes: Based on the measured voltage input and the measured power output at the current moment, determine the induced electromotive force state at the current moment; Based on the induced electromotive force state and discrete time interval at the current moment, the motor slip at the current moment is determined; Based on the current motor slip and electromagnetic state of the induction motor, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state.

[0010] According to the method for identifying equivalent load model parameters provided by the present invention, the noise identification data further includes the system frequency; the frequency characteristic parameter identification based on the noise identification data, the pre-identification result of the equivalent load electromagnetic parameters, and the system frequency, performing parameter regularization, to obtain the effective identification result of the electromagnetic parameters includes: Based on the measured voltage input, measured power output and system frequency at the current moment, the induced electromotive force state at the next moment is predicted to obtain the second predicted induced electromotive force state. Based on the second predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the second predicted power. The error between the effective identification result of electromagnetic parameters and the pre-identification result of electromagnetic parameters of equivalent load is used as the regularization term of the second objective function. Based on the measured power, the second predicted power and the regularization term at the same time, the second objective function is determined. The second objective function is optimized using the differential evolution algorithm to obtain the effective identification results of electromagnetic parameters.

[0011] According to the method for identifying equivalent load model parameters provided by the present invention, the step of identifying the variable torque load inertia based on noise-like identification data, the effective identification results of the electromagnetic parameters, and the electromechanical state of the motor to obtain the motor inertia parameter identification results includes: Based on the measured voltage input, measured power output, and the effective identification results of the electromagnetic parameters at the current moment, the estimated value of the motor slip is determined; Based on the electromechanical state of the motor, determine the predicted value of the motor slip; Based on the motor slip estimate and motor slip prediction at the same time, a third objective function is determined; the objective of the third objective function is to minimize the error between the motor slip prediction and the motor slip estimate. The third objective function is optimized using the differential evolution algorithm to obtain the identification results of the motor inertia parameters.

[0012] According to the method for identifying parameters of an equivalent load model provided by the present invention, the construction of an equivalent load model considering frequency characteristics includes: Based on the system frequency variation, a static constant impedance load model considering frequency characteristics is constructed. The original parameters of the dynamic induction motor load model are transformed, and the dynamic induction motor load model is constructed based on the parameters to be identified obtained by the parameter transformation. Construct an equivalent load model equation that combines the static constant impedance load model and the dynamic induction motor load model in parallel, taking into account frequency characteristics.

[0013] The present invention also provides an identification device for equivalent load model parameters, comprising: The building block is used to construct an equivalent load model that takes into account frequency characteristics; The selection module is used to select noise identification data based on the structure of the equivalent load model. The grading module is used to grade each parameter to be identified according to its identification sensitivity, and to determine the identifiability of each parameter to be identified. The identification module is used to perform step-by-step identification of the equivalent load model parameters based on the noise identification data and the identifiability of each parameter to be identified, including pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification result of the equivalent load model parameters.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the identification method for equivalent load model parameters as described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for identifying equivalent load model parameters as described above.

[0016] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method for identifying equivalent load model parameters as described above.

[0017] The present invention provides a method, apparatus, device, medium, and product for identifying equivalent load model parameters. Considering the need for load frequency-related parameters in the analysis of power consumption characteristics in communication scenarios, this invention constructs an equivalent load model structure that considers frequency characteristics, effectively describing the comprehensive dynamic characteristics of voltage / frequency of electrical equipment. Furthermore, considering the low identifiability of load frequency-related parameters and the increased difficulty in identification due to the increased input quantity and parameters to be identified in the equivalent load model considering frequency characteristics, this invention classifies each parameter to be identified based on identification sensitivity. Based on noise-like identification data and the identifiability of the parameters to be identified, the equivalent load model parameters undergo a step-by-step decoupled identification process: pre-identification, parameter regularization identification, and variable torque load inertia identification. Stable parameter identification under multi-input data is achieved through load frequency characteristic parameter identification with parameter regularization. Variable torque load inertia identification adapts to various mechanical characteristics of real motor loads, thereby obtaining stable and reliable identification results and effectively improving the identifiability of multiple parameters of the equivalent load model under noise-like identification data. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the method for identifying equivalent load model parameters provided in this embodiment of the invention.

[0020] Figure 2 This is a schematic diagram of an application scenario provided by an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of the equivalent load model considering frequency characteristics provided in an embodiment of the present invention.

[0022] Figure 4 This is a schematic diagram of the step-by-step identification process of the equivalent load model parameters provided in the embodiments of the present invention.

[0023] Figure 5 This is a schematic diagram of the step-by-step multi-stage identification strategy for equivalent load model parameters provided in an embodiment of the present invention.

[0024] Figure 6 This is a schematic diagram of the structure of the device for identifying equivalent load model parameters provided in an embodiment of the present invention.

[0025] Figure 7 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0027] With the continuous development of technology, data centers and industrial parks have adopted various energy-saving and efficiency management technologies, including water-cooled air conditioning systems, variable frequency chiller applications, fresh air systems, and modular uninterruptible power supplies (UPS). From the perspective of the power grid, their electrical loads not only include loads exhibiting pure power / impedance characteristics but also a large number of motor-characteristic loads such as air conditioning compressors and water pumps. Simultaneously, with the integration of flexible resources such as energy storage and distributed renewable energy, data centers / industrial parks, as flexible and adjustable loads, can participate in power grid demand-side response and ancillary services in the future. The safe and stable operation of power grid loads, represented by large data centers and industrial parks, is affected by the voltage and frequency stability of the power system, while their active / reactive characteristics can also participate in the voltage and frequency response processes of the power system. Therefore, analyzing the interaction mechanism between large-scale power consumption scenarios in the information and communication field and the power grid is of great significance for improving power supply stability, communication equipment operation stability, and ensuring energy security. Load modeling and parameter identification of data centers or industrial parks from the power grid side is the main means of analyzing the interaction process between information and communication power consumption scenarios and the power grid.

[0028] The main technical approach for integrated load modeling and identification in power system analysis is to use fault-independent, noise-like data for overall identification. This overall identification method is essentially a curve fitting problem, taking voltage as input and power measurement as output. Under the same voltage input, it uses optimization methods to obtain model parameters that minimize the deviation between the measured output and the model's predicted output. Therefore, the equivalent load model in existing technical solutions mainly describes the mathematical relationship between bus voltage and the corresponding load node power. However, when further considering the relationship between load node power, bus voltage, and frequency, the input measurement data in the overall identification method increases, and the set of parameters to be identified grows, leading to low sensitivity and poor identifiability for certain parameters.

[0029] To address the aforementioned issues, this invention, considering the need for load frequency-related parameters in the analysis of power consumption characteristics in communication scenarios, constructs an equivalent load model structure that considers frequency characteristics and involves a static constant impedance load and a dynamic induction motor load connected in parallel. This effectively describes the comprehensive dynamic characteristics of the voltage / frequency of the electrical equipment. Furthermore, considering the low identifiability of load frequency-related parameters and the increased difficulty in identification due to the increase in identification input and parameters after considering frequency characteristics, this invention performs step-by-step identification of the equivalent load model parameters based on noise-like identification data. This decouples the comprehensive load voltage / frequency characteristics and electromagnetic / electromechanical dynamic processes, thereby effectively improving the identifiability of frequency-related parameters.

[0030] Figure 1 This is a flowchart illustrating the method for identifying equivalent load model parameters provided in an embodiment of the present invention. (Refer to...) Figure 1 This invention provides a method for identifying parameters of an equivalent load model, which may specifically include the following steps: Step 110: Construct an equivalent load model that takes into account frequency characteristics.

[0031] It should be noted that the execution subject of the method for identifying equivalent load model parameters provided in this embodiment of the invention can be an electronic device, a component in the electronic device, an integrated circuit, or a chip. The electronic device can be a mobile electronic device or a non-mobile electronic device. For example, a mobile electronic device can be a mobile phone, tablet computer, laptop computer, PDA, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc., while a non-mobile electronic device can be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. This embodiment of the invention does not specifically limit the specific implementation of these methods.

[0032] Please refer to Figure 2 , Figure 2 This is a schematic diagram of an application scenario provided by an embodiment of the present invention. In this embodiment, for scenarios with diversified power equipment in large data centers and industrial parks, in order to ensure the stability of information and communication equipment operation, improve power supply reliability, and ensure information and energy security, equivalent modeling of the load characteristics of the information and communication scenario can be performed from the power grid side to analyze its interaction mechanism with the power system. The power equipment in data centers and industrial parks may include communication equipment, air conditioning equipment, and energy storage systems, etc.

[0033] Please refer to Figure 3 , Figure 3 This is a schematic diagram of the equivalent load model considering frequency characteristics provided in an embodiment of the present invention. In this embodiment, a combined Z+IM load structure of a static constant impedance load Z parallel induction motor (IM) can be selected to construct an equivalent load model considering frequency characteristics, thereby describing the relationship between voltage, frequency and power of the load node through a complete equivalent load model.

[0034] Step 120: Based on the structure of the equivalent load model, select noise identification data.

[0035] Among them, noise identification data can refer to noise-like data used to identify parameters of equivalent load models.

[0036] Based on the structure of the equivalent load model considering frequency characteristics, the data required to identify all equivalent load model parameters can include input data (measured voltage, measured frequency) and output data (active power, reactive power). To ensure that the identification process is independent of fault occurrence, embodiments of the present invention can select noise-like data such as bus voltage, bus frequency, node injected active power, and node injected reactive power under normal operating conditions for the identification of comprehensive equivalent load model parameters.

[0037] Step 130: Classify each parameter to be identified according to its identification sensitivity to determine the identifiability of each parameter to be identified.

[0038] Because the constructed equivalent load model considering frequency characteristics is complete, it has more parameters to be identified and requires more identification data compared to traditional models. Multiple factors contribute to increased identification difficulty, making direct identification impossible. To improve parameter identifiability, this invention can classify multiple parameters based on identification sensitivity. Furthermore, based on the identifiability of the parameters to be identified, a step-by-step decoupled identification process is performed on the equivalent load model parameters: pre-identification, parameter regularization identification, and variable torque load inertia identification. This achieves the identification and acquisition of all parameters of the equivalent load model, thereby effectively improving the identifiability of multi-parameter identification.

[0039] Specifically, multiple parameters to be identified can be classified based on identification sensitivity and parameter characteristics. According to measured data under small fluctuation scenarios during normal load operation, the fluctuation of the selected bus voltage frequency measurement is much smaller than the fluctuation of the voltage measurement. Therefore, the identifiability of voltage-related parameters in the comprehensive equivalent load model can be higher than that of frequency-related parameters. In this embodiment of the invention, the parameters to be identified can be divided into the following categories: Class I: Equivalent load electromagnetic parameters [R] z ,x z,X,X',T d0 This parameter is primarily a voltage-related parameter, exhibiting high identification sensitivity and is also a parameter considered in traditional load model identification. Among them, R... z Let x be the resistance parameter of the constant impedance model. z Here are the reactance parameters for the constant impedance model, where X is the open-circuit reactance, X' is the transient reactance, and T is the reactance. d0 This is the open-circuit time constant.

[0040] Class II: Frequency response parameters [L] dp ,L dq This parameter is also an electromagnetic parameter, but it is a frequency-related parameter, and its identification sensitivity is lower than that of voltage-related parameters. Among them, L... dp and L dq These are the static load active power P s Reactive power Q s The frequency sensitivity coefficient.

[0041] Category III: Motor Inertia Parameters [H] im [A,B], this parameter belongs to the electromechanical parameters of equivalent load, but it is related to frequency characteristics, and its identification sensitivity is lower than that of electromagnetic parameters. Among them, H im Let A be the rotor inertia time constant, A be the coefficient of the quadratic term in the quadratic torque equation, and B be the coefficient of the linear term in the quadratic torque equation.

[0042] In this embodiment of the invention, the identifiability of various parameters can be determined based on their varying sensitivity. To improve parameter identifiability and achieve full identification of the complete parameter set, a corresponding identification algorithm can be designed based on the hierarchical results and features of the parameter set to be identified, thereby enabling the identification and acquisition of all model parameters.

[0043] Step 140: Based on the noise identification data and the identifiability of each parameter to be identified, perform a step-by-step identification process for the equivalent load model parameters, including pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification results of the equivalent load model parameters.

[0044] In this embodiment of the invention, for the set of parameters to be identified in a complete equivalent load model considering frequency characteristics, the electromagnetic parameters related to voltage characteristics and electromagnetic parameters related to frequency characteristics, as well as the electromagnetic parameters and electromechanical parameters, can be decoupled based on noise identification data and the identifiability of each parameter to be identified, so as to achieve step-by-step decoupling identification of equivalent load model parameters: pre-identification of parameters, parameter regularization identification, and identification of load inertia with variable torque.

[0045] By identifying load frequency characteristic parameters through parameter regularization, stable parameter identification can be achieved under multiple input data. By identifying load inertia with variable torque, it is possible to adapt to various mechanical characteristics of real motor loads, thereby effectively improving the identifiability of multi-parameter identification under noise-like identification data.

[0046] This invention addresses the need for load frequency-related parameters in power consumption characteristic analysis of communication scenarios. By constructing an equivalent load model structure that considers frequency characteristics, it can effectively describe the comprehensive dynamic characteristics of voltage / frequency of electrical equipment. Furthermore, considering the low identifiability of load frequency-related parameters and the increased difficulty in identification due to the increased input and parameters to be identified in the equivalent load model with frequency characteristics, this invention classifies each parameter to be identified based on identification sensitivity. Based on the identifiability of noise-like identification data and the parameters to be identified, the equivalent load model parameters are subjected to a step-by-step decoupled identification process: pre-identification, parameter regularization identification, and variable torque load inertia identification. Through parameter regularization of load frequency characteristic parameter identification, stable parameter identification under multi-input data identification is achieved. Through variable torque load inertia identification, various mechanical characteristics of real motor loads are adapted, thereby obtaining stable and reliable identification results and effectively improving the identifiability of multiple parameters of the equivalent load model under noise-like identification data.

[0047] Based on any of the above embodiments, constructing an equivalent load model that considers frequency characteristics may specifically include the following steps: Step 111: Based on the system frequency variation, construct a static constant impedance load model that considers frequency characteristics.

[0048] In this embodiment of the invention, the static charge active power equation can be constructed based on the system frequency change, the frequency sensitivity coefficient of the static load active power, the voltage input, and the resistance parameters; the static charge reactive power equation can be constructed based on the system frequency change, the frequency sensitivity coefficient of the static load reactive power, the voltage input, and the reactance parameters; the static charge active power equation and the static charge reactive power equation can be used to describe the static constant impedance load model considering frequency characteristics.

[0049] Specifically, the static constant impedance load model considering frequency characteristics can be described by the following equation: (1); Where Δω = ω - ω0 is the system frequency change, ω is the system frequency, ω0 is the system synchronous speed, and R... z and x z These are the resistance and reactance parameters of the static constant impedance load model, L. dp and L dq These are the static load active power P s Reactive power Q sThe frequency sensitivity coefficient, the parameter set to be identified for the static constant impedance load model z=[R z ,x z ,L dp ,L dq ].

[0050] Step 112: Perform parameter transformation on the original parameters of the dynamic induction motor load model, and construct the dynamic induction motor load model based on the parameters to be identified obtained from the parameter transformation.

[0051] In this embodiment of the invention, for the dynamic induction motor load model, the original parameters can be transformed, and the dynamic induction motor load model can be constructed based on the parameters to be identified obtained from the parameter transformation, thereby reducing the number of parameters to be identified through parameter transformation. Specifically, the stator reactance parameter x of IM can be transformed. s Rotor reactance parameter x r Magnetizing reactance x m and rotor resistance R r Converted into the following three electromagnetic parameters: open-circuit reactance X, transient reactance X', and open-circuit time constant T. d0 The parameter transformation equation can be as follows: (2); The open-circuit reactance X, transient reactance X', and open-circuit time constant T are obtained through parameter transformation. d0 Once the electromagnetic parameters are obtained, the electromagnetic state equation of the induction motor can be determined: (3); Where ω0 is the system synchronous speed, s σ For motor slip, E d '、E q 'represents the direct-axis d-component and the quadrature-axis q-component of the induced electromotive force vector E', respectively. d I q These are the d-axis and q-axis components of the stator current vector I, respectively.

[0052] Bus voltage vector U = U d +jU q , among which, U d U q These represent the d-axis and q-axis components of the bus voltage vector, respectively, with j being the imaginary unit. In this embodiment of the invention, the relationship between the bus voltage vector U, the induced electromotive force, and the stator current can be determined: (4); The equation of motion for the induction motor rotor can be determined: (5); Among them, H imLet T be the rotor inertia time constant, and T be the electromagnetic torque. e =E d 'I d +E q 'I q T m This refers to mechanical torque.

[0053] In this embodiment of the invention, the mechanical torque T m Rotor speed ω r The functions can include the following three types: constant torque, quadratic torque, and exponential torque: (6); Among them, T m0 For steady-state mechanical torque, ω at steady-state speed r0 Satisfy Aω r0 +Bω r0 +C=1. Considering the general scenario of variable mechanical torque under varying load, T m =f(ω r )T m0 Under the noise-like data identification involved in the embodiments of the present invention, it is possible to determine its steady-state speed ω. r =ω r0 Perform a second-order Taylor expansion nearby: (7); Therefore, in scenarios with small noise fluctuations, a quadratic torque form can be used in this embodiment of the invention to characterize the variable mechanical torque characteristics of any load. Based on the dynamic induction motor load model equation, the set of parameters to be identified for the dynamic induction motor load model is M=[X,X',T]. d0 H im (A, B).

[0054] Step 113: Construct the equivalent load model equations of the parallel connection of the static constant impedance load model and the dynamic induction motor load model, taking into account frequency characteristics.

[0055] In this embodiment of the invention, using the load node voltage amplitude, phase angle, and frequency as inputs, and the node active power and reactive power as outputs, the input-output model equations of the Z+IM integrated equivalent load model considering frequency characteristics can be constructed as follows: (8); Where P is the active power output of the equivalent load model of Z+IM parallel connection, and Q is the reactive power output of the equivalent load model of Z+IM parallel connection.

[0056] Based on any of the above embodiments, the parameters to be identified include motor inertia parameters and electromagnetic parameters with higher identifiability than the motor inertia parameters. The electromagnetic parameters include frequency characteristic parameters and equivalent load electromagnetic parameters with higher identifiability than the frequency characteristic parameters. The step-by-step identification of the equivalent load model parameters based on the noise identification data and the identifiability of each parameter to be identified, to obtain the identification result of the equivalent load model parameters, may specifically include the following steps: Step 141: Based on the noise identification data, perform pre-identification to obtain the pre-identification results of the equivalent load electromagnetic parameters; Step 142: Based on the noise identification data, the pre-identification results of the equivalent load electromagnetic parameters, and the system frequency, frequency characteristic parameter identification with parameter regularization is performed to obtain the effective identification results of the electromagnetic parameters. Step 143: Based on the noise identification data, the effective identification results of the electromagnetic parameters, and the electromechanical state of the motor, the variable torque load inertia is identified to obtain the motor inertia parameter identification results. Step 144: Based on the effective identification results of the electromagnetic parameters and the identification results of the motor inertia parameters, the identification results of the equivalent load model parameters are obtained.

[0057] The equivalent load model proposed in this embodiment of the invention adds frequency measurement to the input quantities, and adds a static load frequency sensitivity coefficient L to the parameters to be identified. dp and L dq Considering the inertial parameter H of variable mechanical torque im The quadratic mechanical torque coefficients A and B present challenges in identification, particularly regarding the poor identifiability of frequency-related parameters. To improve the identifiability of multiple parameters, this invention employs a multi-step identification process—pre-identification, parameter regularization identification, and variable torque load inertia identification—based on the identifiability of the parameter set in a complete equivalent load model considering frequency characteristics. This decouples the load voltage / frequency characteristics and electromagnetic / electromechanical parameters, effectively avoiding the difficulty in identifying certain parameters in noisy data scenarios.

[0058] Please refer to Figure 4 and Figure 5 , Figure 4 This is a schematic diagram of the step-by-step identification process of the equivalent load model parameters provided in the embodiments of the present invention. Figure 5 This is a schematic diagram of the step-by-step multi-stage identification strategy for equivalent load model parameters provided in an embodiment of the present invention.

[0059] After constructing an equivalent load model that considers frequency characteristics, selecting noise identification data, and classifying each parameter to be identified based on identification sensitivity, the equivalent load can be identified in three stages: pre-identification, parameter regularization identification, and load inertia identification with variable torque.

[0060] During identification phase I, frequency characteristics can be disregarded, and pre-identification can be performed based on voltage and power-like noise data to obtain the phase I parameter identification results, which are the equivalent load electromagnetic parameters θ with the highest identifiability. I =[R z,I ,x z,I ,X I ,X I ',T d0,I The "prior value" of the equivalent load electromagnetic parameters is used as the reference value for subsequent identification.

[0061] In identification stage II, the frequency characteristics of the static constant impedance load model can be considered based on identification stage I, using the parameter identification results of stage I as a reference value. This reference value can be used as the regularization term of the objective function, with voltage and power-like noise data as input, and the system frequency added to the input. This allows for the identification of frequency characteristic parameters with parameter regularization, yielding the parameter identification results for stage II, i.e., the effective electromagnetic parameter identification result θ. II =[R z ,x z ,X,X',T d0 ,L dp ,L dq ].

[0062] In stage III of the identification process, noise-like data of voltage and power can be used as input. Based on the effective identification results of electromagnetic parameters from stage II and the estimated electromechanical state of the motor, the variable torque load inertia is identified by combining the rotor motion equation of the induction motor, thus obtaining the stage III parameter identification results, i.e., the motor inertia parameter identification result θ. III =[H im (A,B,C).

[0063] After completing the step-by-step, multi-stage identification of the equivalent load model parameters, the identification result θ=[θ] of the Z+IM integrated equivalent load model parameters considering frequency characteristics can be obtained. II ,θ III ].

[0064] In this embodiment of the invention, multiple identification stages are performed step-by-step and decoupled from each other. Parameter regularization can avoid "overfitting" during the parameter identification process, thereby effectively improving the identifiability of frequency characteristic parameters. Decoupled identification of electromagnetic parameters and electromechanical parameters (i.e., motor inertia parameters) of the induction motor can effectively avoid the problem of poor identifiability caused by direct identification of multiple parameters, and can take into account the characteristics of load-variable mechanical torque, thereby improving the identifiability of multiple parameters of the equivalent load model under noise-like identification data and obtaining stable and reliable identification results.

[0065] Based on any of the above embodiments, the noise identification data includes measured voltage input and measured power output; the pre-identification based on the noise identification data to obtain the pre-identification result of the equivalent load electromagnetic parameters may specifically include the following steps: Step 1411: Based on the measured voltage input and measured power output at the current moment, predict the induced electromotive force state at the next moment to obtain the first predicted induced electromotive force state. Step 1412: Based on the first predicted induced electromotive force state and the measured voltage input at the next moment, predict the power output at the next moment to obtain the first predicted power. Step 1413: Based on the measured power and the first predicted power at the same time, determine the first objective function; the objective of the first objective function is to minimize the deviation between the measured power and the first predicted power. Step 1414: The first objective function is optimized using the differential evolution algorithm to obtain the pre-identification results of the electromagnetic parameters of the equivalent load.

[0066] Continue to refer to Figure 5 In identification stage I, the measured voltage and measured power can be used as inputs. Based on the electromotive force state equation, the induced electromotive force state quantity is predicted. A first predicted power is calculated based on the first predicted induced electromotive force state, and a first objective function is constructed based on the measured power and the first predicted power. The first objective function is then optimized using a differential evolution algorithm. When the deviation between the measured power and the first predicted power is less than a first deviation threshold, the pre-identification result θ of the equivalent load electromagnetic parameters in identification stage I can be obtained. I =[R z,I ,x z,I ,X I ,X I ',T d0,I ].

[0067] In this embodiment of the invention, the identification stage I does not consider frequency characteristics, and the identification method is the same as that for voltage characteristic parameter identification, which can obtain a set of relatively accurate pre-identification parameter results. Specifically, the induced electromotive force state at time k+1 can be predicted using the measured voltage input U and the measured power outputs P and Q at time k+1, thus obtaining the first predicted induced electromotive force state E'. d_p (k+1).

[0068] After obtaining the first predicted induced electromotive force state E' d_p After (k+1), the first predicted induced electromotive force state E' at time k+1 can be used as a basis. d_p (k+1) and the measured voltage input U(k+1), the power output at time k+1 is predicted according to equation (8), and the first predicted power is obtained. The first predicted power includes the first predicted active power P. p(k+1) and the first predicted reactive power Q p (k+1), frequency dynamics are not considered in the prediction process of this stage. After iterative calculation on the N measurement samples within the identification data window, the first objective function of stage I can be obtained as minimizing the deviation between the measured power output and the first predicted power output: (9); Among them, P p Q(k) represents the first predicted active power at time k, P(k) represents the measured active power at time k, and Q(k) represents the measured active power at time k. p Q(k) is the first predicted reactive power at time k, and Q(k) is the measured reactive power at time k.

[0069] In this embodiment of the invention, a differential evolution algorithm with strong global search capabilities can be used to optimize the first objective function and identify the pre-identification result θ of the equivalent load electromagnetic parameters in stage I. I =[R z,I ,x z,I ,X I ,X I ',T d0,I In the identification process of Phase I, the voltage characteristics of the load model are mainly considered. In general noise-like scenarios, the frequency change amplitude is smaller than the voltage change amplitude. Therefore, the identification parameters obtained under the assumption that the frequency affects the load power are highly reliable.

[0070] Based on any of the above embodiments, the step of predicting the induced electromotive force state at the next moment based on the measured voltage input and measured power output at the current moment to obtain a first predicted induced electromotive force state may specifically include the following steps: Step 14111: Based on the measured voltage input and the measured power output at the current moment, determine the induced electromotive force state at the current moment; Step 14112: Determine the motor slip at the current moment based on the induced electromotive force state and discrete time interval at the current moment; Step 14113: Based on the current motor slip and the electromagnetic state of the induction motor, predict the induced electromotive force state at the next moment to obtain the first predicted induced electromotive force state.

[0071] Specifically, according to equations (4) and (8), the induced electromotive force E'(k) at the current moment can be calculated using the measured voltage input U and the measured power outputs P and Q at time k: (10); Among them, P e (k) and Q e(k) represents the active power and reactive power consumed by IM at time k, which is equal to the total load power minus the static load power. The frequency characteristics of the static load are not considered in this stage. (11); Next, the motor slip at time k under the discrete time interval Δt can be calculated: (12); Finally, based on the electromagnetic state equation (3) of the induction motor, the induced electromotive force state at the next moment can be predicted to obtain the first predicted induced electromotive force state E'. d_p (k+1) and E' q_p (k+1): (13); Based on any of the above embodiments, the noise identification data further includes the system frequency; the frequency characteristic parameter identification based on the noise identification data, the pre-identification result of the equivalent load electromagnetic parameters, and the system frequency, after parameter regularization, to obtain the effective identification result of the electromagnetic parameters, may specifically include the following steps: Step 1421: Based on the measured voltage input, measured power output and system frequency at the current moment, predict the induced electromotive force state at the next moment to obtain the second predicted induced electromotive force state. Step 1422: Based on the second predicted induced electromotive force state and the measured voltage input at the next moment, predict the power output at the next moment to obtain the second predicted power. Step 1423: Using the error between the effective identification result of electromagnetic parameters and the pre-identification result of electromagnetic parameters of equivalent load as the regularization term of the second objective function, the second objective function is determined based on the measured power, the second predicted power and the regularization term at the same time. Step 1424: The second objective function is optimized using the differential evolution algorithm to obtain the effective identification results of electromagnetic parameters.

[0072] Continue to refer to Figure 5 In identification stage II, the measured voltage, measured power, and system frequency can be used as inputs. The second predicted power is calculated considering the static load frequency characteristics. The error between the effective electromagnetic parameter identification result of stage II and the pre-identification result of the equivalent load electromagnetic parameters of stage I is used as the regularization term of the second objective function. Based on the measured power, the second predicted power, and the regularization term, the second objective function is constructed. A differential evolution algorithm is used to optimize the second objective function to obtain the effective electromagnetic parameter identification result θ for stage II. II =[R z ,x z ,X,X',T d0 ,Ldp ,L dq ].

[0073] In this embodiment of the invention, identification stage II considers the frequency characteristics of the static load model based on identification stage I, and uses the pre-identification results of the equivalent load electromagnetic parameters obtained in identification stage I as the benchmark value. A regularization term is added to the objective function so that one or more parameters converge to the expected reliable value.

[0074] In this embodiment of the invention, the identification process of identification stage II is basically the same as that of identification stage I. The main difference is that: First, the input of identification stage II increases the system frequency ω. Therefore, when calculating static load power, the influence of frequency change needs to be taken into account, such that equation (11) becomes: (14); In the second prediction of active power P p (k+1) and the second predicted reactive power Q p When calculating (k+1), the frequency dynamics effect term also needs to be considered.

[0075] Secondly, the second objective function y in the identification stage II II Includes parameterized regular expressions: (15); The second objective function includes the sum of squares of the deviations between the measured power output and the predicted power output, as well as the error between the parameters to be identified in Stage II and the parameters identified in Stage I. μ is a weighting coefficient that can be adjusted according to actual conditions. This embodiment of the invention introduces a regularization term to guide the identification parameters to converge to near the identification results in Stage I, thereby effectively avoiding the problem of parameter overfitting.

[0076] During the identification process, it can be reasonably assumed that the active frequency characteristics and reactive frequency characteristics of the static load are equal. Therefore, Stage II identification can obtain the effective identification values ​​θ of the six electromagnetic parameters of the load. II =[R z ,x z ,X,X',T d0 ,L dp ,L dq ].

[0077] The embodiments of the present invention employ a load frequency characteristic parameter identification method with parameter regularization, which facilitates effective convergence of parameters under multiple input data identification, and the identification results are stable and reliable, thus achieving stable parameter identification under multiple input data identification.

[0078] Based on any of the above embodiments, the step of identifying the variable torque load inertia based on noise identification data, the effective identification results of electromagnetic parameters, and the electromechanical state of the motor to obtain the motor inertia parameter identification results may specifically include the following steps: Step 1431: Based on the measured voltage input, measured power output and the effective identification results of the electromagnetic parameters at the current moment, determine the estimated value of the motor slip. Step 1432: Determine the predicted value of motor slip based on the estimated value of motor slip and the electromechanical state of the motor; Step 1433: Based on the estimated and predicted motor slip values ​​at the same time, determine the third objective function; the objective of the third objective function is to minimize the error between the predicted and estimated motor slip values. Step 1434: The third objective function is optimized using the differential evolution algorithm to obtain the motor inertia parameter identification result.

[0079] Continue to refer to Figure 5 In identification stage III, the measured voltage and power can be used as inputs. Based on the identified electromagnetic parameters, the slip state variable is estimated to obtain the slip estimate. The slip is then predicted iteratively based on the motor electromechanical state equation to obtain the slip prediction. A third objective function is constructed based on the slip estimate and the slip prediction. The differential evolution algorithm is used to optimize the third objective function to obtain the motor inertia parameter identification result θ in identification stage III. III =[H im (A,B,C).

[0080] In this embodiment of the invention, after obtaining the effective identification results of electromagnetic parameters through identification stage II, the induced electromotive force state E'(k) at the current moment can be calculated based on the measured voltage input and measured power output at the current moment; the estimated value s of the motor slip can be obtained based on the electromagnetic state equation of equation (12). σ The estimated value of the motor slip is used as the reference value for Stage III identification. Furthermore, the predicted value of the motor slip s can be iteratively calculated based on the rotor motion equation (i.e., the electromechanical state equation) of equation (5). σ_p : (16); The equations for calculating electromagnetic torque and mechanical torque can be: (17); The third objective function for identification stage III can be constructed by minimizing the error between the slip estimate based on the electromagnetic state equation and the slip prediction based on the electromechanical state equation: (18); The differential evolution algorithm can be used to optimize the third objective function and identify the motor inertia parameters for stage III, namely the load inertia constant and torque coefficient θ. III =[H im (A,B,C).

[0081] The embodiments of the present invention utilize a variable torque load inertia identification method, which is beneficial for adapting to various mechanical characteristics of real motor loads.

[0082] This invention, through a step-by-step identification method, can obtain all practical parameters of the integrated equivalent load model, and can acquire frequency characteristic parameters and inertia parameters that can be used for frequency dynamic analysis. It should be noted that in the step-by-step identification process, the identification of the next stage is based on the identification results of the previous stage. Therefore, the identification error in the preceding stage will further lead to corresponding errors in the identification of subsequent stages. This is particularly evident in the identification of load inertia parameters, because the data referenced in stage III identification process is the state variables in the load dynamic process estimated based on the electromagnetic parameters identified in stage II.

[0083] The embodiments of the present invention, by considering frequency characteristics on the basis of the traditional voltage-dominated equivalent load model, considering frequency characteristic parameters in the static load model, and considering variable torque inertia parameters in the dynamic load model, can construct a more complete equivalent load model structure, which can not only reflect the voltage and frequency characteristics of the load at the same time, but also help to understand the power consumption characteristics of complex communication scenarios.

[0084] This invention improves the identifiability of multi-parameter identification through a step-by-step identification algorithm based on a comprehensive equivalent load model using noise-like data. The multi-step identification method, employing pre-identification, parameter regularization identification, and variable torque load inertia identification based on parameter features, decouples load voltage / frequency characteristics and electromagnetic / electromechanical parameters. This effectively avoids the problem of difficult identification of certain parameters in noise-like data scenarios, enabling online real-time acquisition of all parameters of the comprehensive equivalent load model.

[0085] The identification device for equivalent load model parameters provided by the present invention will be described below. The identification device for equivalent load model parameters described below and the identification method for equivalent load model parameters described above can be referred to in correspondence.

[0086] Figure 6 This is a schematic diagram of the structure of the device for identifying equivalent load model parameters provided in an embodiment of the present invention. (Refer to...) Figure 6 This invention provides an identification device for equivalent load model parameters, which may specifically include the following modules: Module 610 is used to construct an equivalent load model that takes into account frequency characteristics; Module 620 is used to select noise identification data based on the structure of the equivalent load model. The grading module 630 is used to grade each parameter to be identified according to the identification sensitivity, and to determine the identifiability of each parameter to be identified. The identification module 640 is used to perform step-by-step identification of the equivalent load model parameters based on the noise identification data and the identifiability of each parameter to be identified, including pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification result of the equivalent load model parameters.

[0087] Based on any of the above embodiments, the parameters to be identified include motor inertia parameters and electromagnetic parameters with higher identifiability than the motor inertia parameters, wherein the electromagnetic parameters include frequency characteristic parameters and equivalent load electromagnetic parameters with higher identifiability than the frequency characteristic parameters; the identification module 640 is specifically used for: Pre-identification is performed based on noise identification data to obtain the pre-identification results of electromagnetic parameters of equivalent load. Based on noise identification data, the pre-identification results of electromagnetic parameters of the equivalent load, and the frequency characteristic parameter identification with parameter regularization of the system frequency, the effective identification results of electromagnetic parameters are obtained. Based on the noise identification data, the effective identification results of the electromagnetic parameters, and the electromechanical state of the motor, the load inertia of the variable torque is identified, and the motor inertia parameter identification results are obtained. The identification result determination submodule is used to obtain the identification results of the equivalent load model parameters based on the effective identification results of the electromagnetic parameters and the identification results of the motor inertia parameters.

[0088] Based on any of the above embodiments, the noise identification data includes a measured voltage input and a measured power output; the identification module 640 is specifically used for: Based on the measured voltage input and measured power output at the current moment, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state. Based on the first predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the first predicted power. Based on the measured power and the first predicted power at the same moment, a first objective function is determined; the objective of the first objective function is to minimize the deviation between the measured power and the first predicted power. The first objective function is optimized using a differential evolution algorithm to obtain the pre-identification results of the electromagnetic parameters of the equivalent load.

[0089] Based on any of the above embodiments, the identification module 640 is specifically used for: Based on the measured voltage input and the measured power output at the current moment, determine the induced electromotive force state at the current moment; Based on the induced electromotive force state and discrete time interval at the current moment, the motor slip at the current moment is determined; Based on the current motor slip and electromagnetic state of the induction motor, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state.

[0090] Based on any of the above embodiments, the noise identification data further includes system frequency; the identification module 640 is specifically used for: Based on the measured voltage input, measured power output and system frequency at the current moment, the induced electromotive force state at the next moment is predicted to obtain the second predicted induced electromotive force state. Based on the second predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the second predicted power. The error between the effective identification result of electromagnetic parameters and the pre-identification result of electromagnetic parameters of equivalent load is used as the regularization term of the second objective function. Based on the measured power, the second predicted power and the regularization term at the same time, the second objective function is determined. The second objective function is optimized using the differential evolution algorithm to obtain the effective identification results of electromagnetic parameters.

[0091] Based on any of the above embodiments, the identification module 640 is specifically used for: Based on the measured voltage input, measured power output, and the effective identification results of the electromagnetic parameters at the current moment, the estimated value of the motor slip is determined; Based on the electromechanical state of the motor, determine the predicted value of the motor slip; Based on the motor slip estimate and motor slip prediction at the same time, a third objective function is determined; the objective of the third objective function is to minimize the error between the motor slip prediction and the motor slip estimate. The third objective function is optimized using the differential evolution algorithm to obtain the identification results of the motor inertia parameters.

[0092] Based on any of the above embodiments, the construction module 610 is specifically used for: Based on the system frequency variation, a static constant impedance load model considering frequency characteristics is constructed. The original parameters of the dynamic induction motor load model are transformed, and the dynamic induction motor load model is constructed based on the parameters to be identified obtained by the parameter transformation. Construct an equivalent load model equation that combines the static constant impedance load model and the dynamic induction motor load model in parallel, taking into account frequency characteristics.

[0093] This invention addresses the need for load frequency-related parameters in power consumption characteristic analysis of communication scenarios. By constructing an equivalent load model structure that considers frequency characteristics, it can effectively describe the comprehensive dynamic characteristics of voltage / frequency of electrical equipment. Furthermore, considering the low identifiability of load frequency-related parameters and the increased difficulty in identification due to the increased input and parameters to be identified in the equivalent load model with frequency characteristics, this invention classifies each parameter to be identified based on identification sensitivity. Based on the identifiability of noise-like identification data and the parameters to be identified, the equivalent load model parameters are subjected to a step-by-step decoupled identification process: pre-identification, parameter regularization identification, and variable torque load inertia identification. Through parameter regularization of load frequency characteristic parameter identification, stable parameter identification under multi-input data identification is achieved. Through variable torque load inertia identification, various mechanical characteristics of real motor loads are adapted, thereby obtaining stable and reliable identification results and effectively improving the identifiability of multiple parameters of the equivalent load model under noise-like identification data.

[0094] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, communications interface 720, and memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute an identification method for equivalent load model parameters. This method includes: constructing an equivalent load model considering frequency characteristics; selecting noise-type identification data based on the structure of the equivalent load model; classifying each parameter to be identified according to its identification sensitivity to determine the identifiability of each parameter; and performing a step-by-step identification process—pre-identification, parameter regularization identification, and variable torque load inertia identification—on the equivalent load model parameters based on the noise-type identification data and the identifiability of each parameter to be identified, to obtain the identification result of the equivalent load model parameters.

[0095] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the identification method for equivalent load model parameters provided by the above methods. The method includes: constructing an equivalent load model considering frequency characteristics; selecting noise-like identification data based on the structure of the equivalent load model; classifying each parameter to be identified according to its identification sensitivity to determine the identifiability of each parameter to be identified; and performing a step-by-step identification of the equivalent load model parameters based on the noise-like identification data and the identifiability of each parameter to be identified, including pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification result of the equivalent load model parameters.

[0097] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for identifying equivalent load model parameters provided by the methods described above. This method includes: constructing an equivalent load model considering frequency characteristics; selecting noise-like identification data based on the structure of the equivalent load model; classifying each parameter to be identified according to its identification sensitivity to determine the identifiability of each parameter; and performing a step-by-step identification process—pre-identification, parameter regularization identification, and variable torque load inertia identification—on the equivalent load model parameters based on the noise-like identification data and the identifiability of each parameter to be identified, to obtain the identification result of the equivalent load model parameters.

[0098] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0099] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0100] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying parameters of an equivalent load model, characterized in that, include: Construct an equivalent load model that considers frequency characteristics; Based on the structure of the equivalent load model, noise identification data is selected. The identifiability of each parameter to be identified is determined by classifying each parameter based on its identification sensitivity. Based on the noise identification data and the identifiability of each parameter to be identified, the equivalent load model parameters are identified in a step-by-step manner: pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification results of the equivalent load model parameters.

2. The method for identifying equivalent load model parameters according to claim 1, characterized in that, The parameters to be identified include motor inertia parameters and electromagnetic parameters with higher identifiability than the motor inertia parameters. The electromagnetic parameters include frequency characteristic parameters and equivalent load electromagnetic parameters with higher identifiability than the frequency characteristic parameters. The identification process, based on the noise identification data and the identifiability of each parameter to be identified, involves a step-by-step identification process for the equivalent load model parameters: pre-identification, parameter regularization identification, and variable torque load inertia identification. This process yields the identification results for the equivalent load model parameters, including: Pre-identification is performed based on noise identification data to obtain the pre-identification results of electromagnetic parameters of equivalent load. Based on noise identification data, the pre-identification results of electromagnetic parameters of the equivalent load, and the frequency characteristic parameter identification with parameter regularization of the system frequency, the effective identification results of electromagnetic parameters are obtained. Based on the noise identification data, the effective identification results of the electromagnetic parameters, and the electromechanical state of the motor, the load inertia of the variable torque is identified, and the motor inertia parameter identification results are obtained. Based on the effective identification results of the electromagnetic parameters and the identification results of the motor inertia parameters, the identification results of the equivalent load model parameters are obtained.

3. The method for identifying equivalent load model parameters according to claim 2, characterized in that, The noise identification data includes measured voltage input and measured power output; the pre-identification based on the noise identification data to obtain the pre-identification results of the equivalent load electromagnetic parameters includes: Based on the measured voltage input and measured power output at the current moment, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state. Based on the first predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the first predicted power. Based on the measured power and the first predicted power at the same moment, a first objective function is determined; the objective of the first objective function is to minimize the deviation between the measured power and the first predicted power. The first objective function is optimized using a differential evolution algorithm to obtain the pre-identification results of the electromagnetic parameters of the equivalent load.

4. The method for identifying equivalent load model parameters according to claim 3, characterized in that, The prediction of the induced electromotive force state at the next moment based on the measured voltage input and measured power output at the current moment, to obtain the first predicted induced electromotive force state, includes: Based on the measured voltage input and the measured power output at the current moment, determine the induced electromotive force state at the current moment; Based on the induced electromotive force state and discrete time interval at the current moment, the motor slip at the current moment is determined; Based on the current motor slip and electromagnetic state of the induction motor, the induced electromotive force state at the next moment is predicted to obtain the first predicted induced electromotive force state.

5. The method for identifying equivalent load model parameters according to claim 3, characterized in that, The noise identification data also includes the system frequency; the frequency characteristic parameter identification based on the noise identification data, the pre-identification results of the equivalent load electromagnetic parameters, and the system frequency, with parameter regularization, yields effective electromagnetic parameter identification results, including: Based on the measured voltage input, measured power output and system frequency at the current moment, the induced electromotive force state at the next moment is predicted to obtain the second predicted induced electromotive force state. Based on the second predicted induced electromotive force state and the measured voltage input at the next moment, the power output at the next moment is predicted to obtain the second predicted power. The error between the effective identification result of electromagnetic parameters and the pre-identification result of electromagnetic parameters of equivalent load is used as the regularization term of the second objective function. Based on the measured power, the second predicted power and the regularization term at the same time, the second objective function is determined. The second objective function is optimized using the differential evolution algorithm to obtain the effective identification results of electromagnetic parameters.

6. The method for identifying equivalent load model parameters according to claim 2, characterized in that, The variable torque load inertia identification based on noise-type identification data, the effective identification results of electromagnetic parameters, and the electromechanical state of the motor, to obtain the motor inertia parameter identification results, includes: Based on the measured voltage input, measured power output, and the effective identification results of the electromagnetic parameters at the current moment, the estimated value of the motor slip is determined; Based on the electromechanical state of the motor, determine the predicted value of the motor slip; Based on the motor slip estimate and motor slip prediction at the same time, a third objective function is determined; the objective of the third objective function is to minimize the error between the motor slip prediction and the motor slip estimate. The third objective function is optimized using the differential evolution algorithm to obtain the identification results of the motor inertia parameters.

7. The method for identifying equivalent load model parameters according to claim 1, characterized in that, The construction of the equivalent load model considering frequency characteristics includes: Based on the system frequency variation, a static constant impedance load model considering frequency characteristics is constructed. The original parameters of the dynamic induction motor load model are transformed, and the dynamic induction motor load model is constructed based on the parameters to be identified obtained by the parameter transformation. Construct an equivalent load model equation that combines the static constant impedance load model and the dynamic induction motor load model in parallel, taking into account frequency characteristics.

8. A device for identifying parameters of an equivalent load model, characterized in that, include: The building block is used to construct an equivalent load model that takes into account frequency characteristics; The selection module is used to select noise identification data based on the structure of the equivalent load model. The grading module is used to grade each parameter to be identified according to its identification sensitivity, and to determine the identifiability of each parameter to be identified. The identification module is used to perform step-by-step identification of the equivalent load model parameters based on the noise identification data and the identifiability of each parameter to be identified, including pre-identification, parameter regularization identification, and variable torque load inertia identification, to obtain the identification result of the equivalent load model parameters.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for identifying the equivalent load model parameters as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for identifying the equivalent load model parameters as described in any one of claims 1 to 7.

11. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for identifying the equivalent load model parameters as described in any one of claims 1 to 7.

Images