Load identification-oriented detection signal two-stage optimization method and device

Abstract

The invention relates to the technical field of load identification, in particular to a load identification-oriented detection signal two-stage optimization method and device, and the method comprises the steps: determining a discrete frequency set of a frequency optimization range, so as to calculate and store a partial derivative matrix, an initial detection signal evaluation and optimization model for load model identification is constructed according to the partial derivative matrix; introducing a frequency penalty term into the model to obtain an optimization model oriented to frequency optimization, and solving the optimization model oriented to frequency optimization to obtain a frequency optimization result; and defining each frequency point of the model by using the frequency optimization result to obtain an optimization model oriented to amplitude optimization, and solving the optimization model oriented to amplitude optimization to obtain an amplitude optimization result. Therefore, the problems that in related optimization problems, the frequency variable remarkably influences the influence of the amplitude variable on the target function value, the partial derivative of the input and output function of the load model needs to be frequently calculated in the optimization process, and the optimization speed is seriously restricted are solved.

Description

Technical Field

[0001] This invention relates to the field of load identification technology, and in particular to a two-stage optimization method and apparatus for detection signals for load identification. Background Technology

[0002] Load modeling based on probe signals is an emerging load modeling technology. Its core lies in actively injecting specific probe signals into the power grid through controllable devices to stimulate desired load dynamic responses, which are then used for load parameter identification. Compared to traditional load modeling methods, load modeling based on probe signals offers greater operability due to the controllability of the probe signal strength and waveform. By appropriately adjusting the properties of the probe signals, suitable load dynamic responses can be stimulated more effectively while ensuring system safety and stability, thereby improving the accuracy of load modeling results.

[0003] Since the probe signal directly determines the dynamic characteristics of the excited load, the effectiveness of load modeling based on the probe signal largely depends on the properties of the injected probe signal. Therefore, designing a high-quality probe signal is key to improving this method. The design of the probe signal mainly involves its waveform, frequency distribution, and amplitude. Even after the basic waveform of the signal is determined, its spectral characteristics and amplitude distribution must still be considered comprehensively. In previous studies, related technologies have established a theoretical model for load parameter identification errors, and based on this model, a probe signal evaluation and optimization model for load model identification has been constructed, as shown below:

[0004] in, For a set of frequencies, For the set of amplitudes of the corresponding frequency components, and To optimize the model's optimization objective, The weight matrix is ​​the parameter. Var Represents variance. The deviation in the identification of load model parameters The Hadamard product is a direct product of corresponding elements at the same position in a matrix of the same dimension. This is a noise characteristic distribution vector, describing the distribution of noise across different frequency bands. m To find the number of frequencies, To detect the upper limit of the total energy of the signal, and To constrain the upper and lower limits of the frequency of the detection signal. To constrain the upper limit of the amplitude of each frequency component of the probe signal, H is the input signal action matrix, which describes the influence of the input signal on the parameter identification results under a specific load model structure, and is defined as follows:

[0005]

[0006] in, K and Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model.

[0007] In optimization problems, both frequency and amplitude are optimization variables, and there is a close coupling between them. The frequency variable significantly affects the influence of the amplitude variable on the objective function value, making it difficult to optimize both types of variables simultaneously. Furthermore, the optimization process requires frequent calculations of the partial derivatives of the load model's input and output functions, an operation with enormous computational cost that severely restricts optimization speed. Therefore, a more efficient optimization process is urgently needed to significantly improve computational efficiency while maintaining optimization accuracy as much as possible. Summary of the Invention

[0008] This invention provides a two-stage optimization method and apparatus for detection signals for load identification, which addresses the problems in related optimization problems where frequency variables significantly affect the amplitude variables on the objective function value, and where frequent calculations of the partial derivatives of the load model's input and output functions are required during the optimization process, severely restricting the optimization speed.

[0009] A first aspect of the present invention provides a two-stage optimization method for detection signals for load identification, comprising the following steps: A discrete frequency set is determined for the frequency optimization range, and the partial derivative matrix is ​​calculated and stored. Based on the partial derivative matrix, an initial probe signal evaluation and optimization model for load model identification is constructed. A frequency penalty term is introduced into the initial detection signal evaluation and optimization model to obtain a frequency-optimized optimization model, and the frequency-optimized optimization model is solved to obtain the frequency optimization result. The frequency points of the initial detection signal evaluation and optimization model are defined using the frequency optimization results to obtain an amplitude optimization model. The amplitude optimization model is then solved to obtain the amplitude optimization results.

[0010] Optionally, the specific expression for the partial derivative matrix is:

[0011] in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

[0012] Optionally, solving the frequency-optimized model to obtain the frequency optimization result includes: The decision variables of the frequency optimization model are randomly initialized within the range of their values ​​to obtain the first type group; Three individuals are randomly selected from the first population to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, a random number between 0 and 1 is generated. If the random number is less than the first preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target frequency function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population described above is executed iteratively until the target frequency function value converges, thus obtaining the frequency optimization result.

[0013] Optionally, solving the magnitude-oriented optimization model to obtain the magnitude optimization result includes: The decision variables of the optimization model oriented towards magnitude optimization are randomly initialized within the range of their values ​​to obtain the second population; Three individuals are randomly selected from the second population to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, a random number between 0 and 1 is generated. If the random number is less than the second preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target magnitude function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population described above is executed iteratively until the target magnitude function value converges, thus obtaining the magnitude optimization result.

[0014] A second aspect of the present invention provides a two-stage optimization device for detection signals oriented towards load identification, comprising: An initial construction module is used to determine the set of discrete frequencies for frequency optimization, to calculate and store the partial derivative matrix, and to construct an initial probe signal evaluation and optimization model for load model identification based on the partial derivative matrix. The frequency optimization module is used to introduce a frequency penalty term into the initial detection signal evaluation and optimization model to obtain a frequency-optimized optimization model, and to solve the frequency-optimized optimization model to obtain the frequency optimization result. The amplitude optimization module is used to limit the frequency points of the initial detection signal evaluation and optimization model using the frequency optimization results, so as to obtain an amplitude optimization model, and solve the amplitude optimization model to obtain the amplitude optimization result.

[0015] Optionally, the specific expression for the partial derivative matrix is:

[0016] in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

[0017] Optionally, the frequency optimization module includes: The first initialization unit is used to randomly initialize the decision variables of the optimization model oriented towards frequency optimization within the range of their values, so as to obtain the first type group. The first generation unit is used to randomly select three individuals from the first population and iterate N times to generate the next generation of individuals, where N is a positive integer; The first crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the first preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The first evolutionary unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target frequency function value for the next generation of evolution; The first parallel unit is used to execute the generation, crossover, and selection processes a preset number of times in parallel to obtain the next generation population; The first iteration unit is used to iteratively execute the aforementioned process of generating the next generation population until the target frequency function value converges, thereby obtaining the frequency optimization result.

[0018] Optionally, the amplitude optimization module includes: The second initialization unit is used to randomly initialize the decision variables within the range of values ​​of the magnitude of the optimization model oriented towards magnitude optimization, so as to obtain the second population. The second generation unit is used to randomly select three individuals from the second population and iterate N times to generate the next generation of individuals, where N is a positive integer; The second crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the second preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The second evolutionary unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target amplitude function value for the next generation of evolution; The second parallel unit is used to execute the generation, crossover, and selection processes a preset number of times in parallel to obtain the next generation population; The second iteration unit is used to iteratively execute the aforementioned process of generating the next generation population until the target magnitude function value converges, thereby obtaining the magnitude optimization result.

[0019] A third aspect of the present invention 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 program to implement the two-stage optimization method for load identification detection signals as described in the above embodiments.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described two-stage optimization method for load-oriented detection signals.

[0021] The two-stage optimization method and apparatus for detection signals oriented to load identification proposed in this invention effectively reduces the complexity of the optimization problem while maintaining basic computational accuracy by employing a discrete frequency set and frequency-amplitude two-stage optimization method. This improves the efficiency of optimization calculation and the success rate of optimization search, providing a more feasible and efficient solution for the calculation of detection signal optimization problems oriented to load identification.

[0022] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0023] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart of a two-stage optimization method for detection signals for load identification according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the specific execution of a two-stage optimization method for detection signals for load identification according to an embodiment of the present invention. Figure 3 This is a schematic diagram illustrating the descent of the objective function during an iterative process according to an embodiment of the present invention; Figure 4 This is a block diagram of a two-stage optimization device for detection signals for load identification according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention.

[0024] Explanation of reference numerals in the attached figures: 40-Two-stage optimization device for detection signals for load identification, 401-Initial construction module, 402-Frequency optimization module, 403-Amplitude optimization module, 501-Memory, 502-Processor and 503-Communication interface. Detailed Implementation

[0025] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0026] The following describes, with reference to the accompanying drawings, a two-stage optimization method and apparatus for load identification detection signals according to embodiments of the present invention.

[0027] Figure 1This is a flowchart illustrating a two-stage optimization method for detection signals oriented towards load identification, provided in an embodiment of the present invention.

[0028] like Figure 1 As shown, the two-stage optimization method for detection signals oriented towards load identification includes the following steps: In step S101, a set of discrete frequencies within the frequency optimization range is determined to calculate and store the partial derivative matrix, and an initial probe signal evaluation and optimization model for load model identification is constructed based on the partial derivative matrix.

[0029] In actual implementation, the calculation of the partial derivative matrix J of the load model input and output channel functions is the most complex, because it involves the calculation of a large number of partial derivatives of the load model channel functions.

[0030]

[0031] In the formula, For the first m angular frequency, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter.

[0032] However, the J matrix is ​​essentially composed of several frequency-based partial derivative submatrices. That is:

[0033] in:

[0034] at the same time, and All are about A continuous function, therefore, if two frequencies If they are close to each other, then they are continuous functions. and ,as well as and They will be very close. Given the submatrix yes and Therefore, the submatrices J corresponding to the two frequencies are very similar. From this perspective, optimization within the set of discrete frequencies is feasible. This is because, when the frequency resolution is set appropriately, the matrix J corresponding to the optimal frequency and its neighboring discrete frequencies is highly similar, and thus their impact on the objective function is also similar. Therefore, optimization on discrete frequencies does not significantly affect the optimization result.

[0035] Therefore, embodiments of the present invention propose selecting the frequency of the optimized detection signal from a discrete set, namely:

[0036] in, To optimize the frequency set, Let the set of discrete frequencies be defined as follows:

[0037] in, For frequency resolution, M is the number of frequencies in the discrete frequency set. Because the discrete frequency set... To determine the set, the partial derivative matrix J can be calculated and stored in advance. large :

[0038] in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

[0039] The advantage of this pre-calculation and recording method is that it significantly reduces the computational complexity of the optimization process. In iterative optimization of signal design, parameters need to be constantly updated, and each update requires recalculating the objective function. Therefore, it is necessary to repeatedly calculate different... The corresponding partial derivative matrix J. Because of the optimization... Belongs to the discrete frequency set A subset of the partial derivative matrix J, which has been pre-computed and stored. large Under the premise that, it can be extracted directly from The corresponding submatrix J. This approach eliminates the need for repeated calculations of the load model's input and output functions. and The steps for calculating their partial derivatives effectively reduce the single-time computation cost of the objective function and improve the overall optimization efficiency.

[0040] In step S102, a frequency penalty term is introduced into the initial detection signal evaluation and optimization model to obtain a frequency-optimized optimization model, and the frequency-optimized optimization model is solved to obtain the frequency optimization result.

[0041] In step S103, the frequency points of the initial detection signal evaluation and optimization model are limited by the frequency optimization results to obtain the amplitude optimization model, and the amplitude optimization model is solved to obtain the amplitude optimization result.

[0042] In some embodiments, the frequency optimization model is solved to obtain frequency optimization results, including: The decision variables of the frequency optimization model are randomly initialized within the range of their values ​​to obtain the first type group. Three individuals are randomly selected from the first group to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, generate a random number between 0 and 1. If the random number is less than the first preset crossover probability, perform crossover replacement; otherwise, retain the original individual elements and finally obtain the crossover individual. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target frequency function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population is executed iteratively until the target frequency function value converges, and the frequency optimization result is obtained.

[0043] In some embodiments, the optimization model oriented towards magnitude optimization is solved to obtain magnitude optimization results, including: The decision variables of the optimization model oriented towards magnitude optimization are randomly initialized within the range of their values ​​to obtain the second group; Three individuals are randomly selected from the second population to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, generate a random number between 0 and 1. If the random number is less than the second preset crossover probability, perform crossover replacement; otherwise, retain the original individual elements and finally obtain the crossover individual. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target magnitude function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population is executed iteratively until the target magnitude function value converges, and the magnitude optimization result is obtained.

[0044] In practical implementation, besides calculating the partial derivative matrix, the simultaneous optimization of frequency and amplitude decision variables is also a challenge. Since the amplitude decision variable corresponds one-to-one with the frequency variable, a change in frequency will significantly alter its impact on the objective function, even if the amplitude variable remains constant. Furthermore, considering that the frequency of the excitation signal directly determines the main characteristics of the excited load dynamics, determining a suitable core frequency can effectively improve the identifiability of the excited load dynamics, thereby improving the accuracy of parameter identification. Based on the above analysis, this invention proposes a two-stage detection signal frequency-amplitude optimization method. This method decomposes the originally coupled frequency and amplitude joint optimization problem into two sequentially executed steps: first determining the frequency, then optimizing the amplitude. This step-by-step strategy reduces the overall optimization complexity.

[0045] Specifically, the first step is to select core frequencies to determine those that have a critical impact on the identification results. Assuming that the amplitudes of all frequencies remain consistent and the total energy meets the energy upper limit, the optimization model (i.e., the frequency-oriented optimization model) is as follows:

[0046] in, For a set of frequencies, Let H be the weight matrix of the parameters, and let H be the input signal effect matrix. This matrix describes how the input signal influences the parameter identification results under a specific load model structure. The Hadamard product is a direct product of corresponding elements at the same position in a matrix of the same dimension. Var Represents variance. This is a noise characteristic distribution vector, describing the distribution of noise across different frequency bands. For frequency term penalty coefficient, This is the frequency interval threshold. To constrain the upper limit of the frequency of the detected signal, m To find the number of frequencies, To determine the upper limit of the total energy of the detected signal, two key improvements were introduced in the first stage of optimization based on the original optimization model: firstly, the frequency optimization range was limited to the set of discrete frequencies. Secondly, a frequency penalty term is added. The main reason for introducing this penalty term is that the original model optimization result may result in frequency intervals that are too small or even repetitive; however, in the second stage of optimization, the amplitude will be redistributed, and its effect is somewhat similar to frequency repetition. Therefore, frequency repetition in the first stage optimization result is not the final desired outcome, as it not only does not help improve the recognition effect but may also reduce the diversity of the signal. To this end, this embodiment of the invention introduces a frequency penalty term into the model, which is applied when the interval between any two frequencies is less than a threshold. When the frequency is too concentrated, the objective function will have an additional penalty coefficient term, which effectively avoids the situation where the frequency is too concentrated.

[0047] Furthermore, considering the complex nonlinear characteristics of the proposed model, traditional gradient-based algorithms are difficult to solve effectively. Therefore, in this embodiment of the invention, the differential evolution algorithm in global optimization algorithms is used to solve the frequency-oriented optimization model, thereby obtaining the frequency optimization result. The solution process is as follows: (1) Population initialization: within the range of values ​​of the decision variable frequencies Initialize randomly within the population to generate a population with the following number of individuals: N The decision variable dimension is m .

[0048]

[0049] in, Refers to the initial first i The first individual j Each component.

[0050] (2) Mutation: Three individuals are randomly selected from the previous generation to generate the next generation of individuals, where N is a positive integer, i.e.:

[0051] in, and , The result obtained after the Gth iteration is shown. i Individual vectors, Represents the G+1th mutation. i Individual vectors.

[0052] (3) Crossover: To avoid the generated individuals during the iteration process becoming merely a linear combination of the initial population, a crossover operation is introduced. This operation enhances population diversity by replacing some elements in the current individual with elements at corresponding positions in the previous generation. Specifically, a random number between 0 and 1 is generated for each element in the individual; if the random number is less than the preset crossover probability, crossover is performed; otherwise, the original individual elements are retained.

[0053] in, The first after the cross G+ The first generation i The first individual j One portion, This represents the crossover rate.

[0054] (4) Selection: To ensure the objective function decreases, individuals after crossover... To connect with the previous generation of individuals Compare the results. The smaller objective function value is used for the next generation of evolution.

[0055] in, f Refers to the objective function; if no constraints are violated, the objective function is... Otherwise, the objective function value is Inf, i.e., infinity. When calculating the objective function, based on the aforementioned points, the partial derivative matrix J can be directly obtained from the stored partial derivative matrix J. large Extracting the structure from the data reduces the computational load.

[0056] Each population must sequentially undergo the aforementioned mutation, crossover, and selection operations, and these operations must be performed in parallel a predetermined number of times to preserve superior individuals for the next generation. After obtaining the next generation, the process of generating the next generation is iteratively executed. Through this mechanism, the target frequency function value can be gradually reduced, and the optimization result continuously approaches the global optimum, thus completing the entire differential evolution process and obtaining the final frequency optimization result. .

[0057] Furthermore, building upon the first stage of frequency optimization, the second step will optimize the amplitudes corresponding to each frequency component in the probe signal. This stage will directly utilize the frequency results determined in the first stage. This process limits the optimization variables to the amplitude at each frequency point. The main objective of this step is to redistribute the amplitude energy at each frequency point while satisfying the total signal energy constraint, in order to maximize the effectiveness of the probe signal in load parameter identification. The optimization model obtained from this process (i.e., the optimization model oriented towards amplitude optimization) is shown below:

[0058] in, yes The first in i Each element.

[0059] Furthermore, the differential evolution algorithm is used to solve the optimization model oriented towards magnitude optimization. Similarly, through population initialization, mutation, crossover, and selection, the entire differential evolution process is completed, yielding the final optimization result. .

[0060] The following detailed description of the two-stage optimization method for detection signals for load identification proposed in this invention will be provided through a specific embodiment.

[0061] This specific embodiment compares the effects of solving the detection signal optimization model under three different conditions: (1) Test 1: using a continuous frequency set, performing 50 optimization tests; (2) Test 2: using a discrete frequency set, optimizing both frequency and amplitude simultaneously, performing 100 optimization tests; (3) Test 3: using a discrete frequency set, optimizing both frequency and amplitude separately, performing 100 optimization tests. The minimum value of the objective function obtained by optimization, the average computation time, and the optimization effect are shown in Table 1 below. Meanwhile, the decrease of the objective function with iterative optimization under the three conditions is shown in Figure 3.

[0062] Table 1 Comparison of Optimization Algorithm Efficiency

[0063] The comparative results show that when using a continuous frequency set for optimization, the computational cost increases significantly due to the need to repeatedly calculate the load model's input and output functions and their partial derivatives, resulting in a much longer computation time compared to using a discrete frequency set. Furthermore, the two-stage optimization method proposed in this embodiment of the invention has a computation time that is only about 70% of that of synchronous frequency-amplitude optimization, demonstrating the effectiveness of this method in improving computational efficiency.

[0064] From the final obtained objective function value ( Obj As can be seen from the results, the results in all three cases are far lower than the initial solution, and the numerical values ​​are quite close, indicating that the method of using discrete frequency sets and staged optimization can still maintain good computational accuracy. Although the frequency-amplitude synchronous optimization theoretically has a smaller lower bound on the objective function, the solution process is less stable due to the strong nonlinearity of this optimization problem. Table 1 shows that the synchronous optimization method has only a 34% probability of obtaining an objective function value lower than that of the two-stage optimization method, while the two-stage optimization method has a 63% probability of converging to its optimal solution.

[0065] In summary, the two-stage optimization method for detection signals oriented towards load identification proposed in this embodiment of the invention effectively reduces the complexity of the optimization problem while maintaining basic computational accuracy by employing a discrete frequency set and a frequency-amplitude two-stage optimization approach. This improves the efficiency of optimization calculation and the success rate of finding the optimal solution, providing a more feasible and efficient solution for the calculation of detection signal optimization problems oriented towards load identification.

[0066] Next, referring to the accompanying drawings, a two-stage optimization device for detection signals oriented towards load identification according to an embodiment of the present invention is described.

[0067] Figure 4 This is a block diagram of a two-stage optimization device for detection signals oriented towards load identification, provided in an embodiment of the present invention.

[0068] like Figure 4 As shown, the two-stage optimization device 40 for load identification detection signals includes: an initial construction module 401, a frequency optimization module 402, and an amplitude optimization module 403.

[0069] The initial construction module 401 is used to determine the discrete frequency set of the frequency optimization range, calculate and store the partial derivative matrix, and construct an initial probe signal evaluation and optimization model for load model identification based on the partial derivative matrix. The frequency optimization module 402 introduces a frequency penalty term into the initial probe signal evaluation and optimization model to obtain a frequency-optimized optimization model, and solves the frequency-optimized optimization model to obtain the frequency optimization result. The amplitude optimization module 403 uses the frequency optimization result to limit each frequency point of the initial probe signal evaluation and optimization model to obtain an amplitude-optimized optimization model, and solves the amplitude-optimized optimization model to obtain the amplitude optimization result.

[0070] In some embodiments, the specific expression for the partial derivative matrix is:

[0071] in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

[0072] In some embodiments, the frequency optimization module 402 includes: The first initialization unit is used to randomly initialize the decision variables within the range of values ​​of the frequency of the optimization model oriented towards frequency optimization, so as to obtain the first type group. The first generation unit is used to randomly select three individuals from the first group to generate the next generation of individuals, where N is a positive integer; The first crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the first preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The first iteration unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target frequency function value for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population is executed iteratively until the target frequency function value converges, and the frequency optimization result is obtained.

[0073] In some embodiments, the amplitude optimization module 403 includes: The second initialization unit is used to randomly initialize the decision variables within the range of values ​​of the decision variables in the optimization model oriented towards magnitude optimization, so as to obtain the second population. The second generation unit is used to randomly select three individuals from the second population to generate the next generation of individuals, where N is a positive integer. The second crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the second preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The second iteration unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target magnitude function value for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population is executed iteratively until the target magnitude function value converges, and the magnitude optimization result is obtained.

[0074] It should be noted that the explanation of the above-mentioned two-stage optimization method for load identification detection signals also applies to the two-stage optimization device for load identification detection signals in this embodiment, and will not be repeated here.

[0075] The two-stage optimization device for detection signals oriented to load identification proposed in this embodiment of the invention effectively reduces the complexity of the optimization problem while maintaining the basic calculation accuracy by adopting a two-stage optimization method of discrete frequency set and frequency-amplitude. This improves the efficiency of optimization calculation and the success rate of optimization search, providing a more feasible and efficient solution for the calculation of detection signal optimization problems oriented to load identification.

[0076] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0077] The electronic device may include: The memory 501, the processor 502, and the computer program stored on the memory 501 and capable of running on the processor 502.

[0078] When the processor 502 executes the program, it implements the two-stage optimization method for detection signals for load identification provided in the above embodiments.

[0079] Furthermore, electronic devices also include: Communication interface 503 is used for communication between memory 501 and processor 502.

[0080] The memory 501 is used to store computer programs that can run on the processor 502.

[0081] Memory 501 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0082] If the memory 501, processor 502, and communication interface 503 are implemented independently, then the communication interface 503, memory 501, and processor 502 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0083] Optionally, in a specific implementation, if the memory 501, processor 502, and communication interface 503 are integrated on a single chip, then the memory 501, processor 502, and communication interface 503 can communicate with each other through an internal interface.

[0084] Processor 502 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0085] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described two-stage optimization method for load-oriented detection signals.

[0086] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0087] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0088] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0089] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0090] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0091] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0092] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0093] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A two-stage optimization method for detection signals for load identification, characterized in that, Includes the following steps: A discrete frequency set is determined for the frequency optimization range, and the partial derivative matrix is ​​calculated and stored. Based on the partial derivative matrix, an initial probe signal evaluation and optimization model for load model identification is constructed. A frequency penalty term is introduced into the initial detection signal evaluation and optimization model to obtain a frequency-optimized optimization model, and the frequency-optimized optimization model is solved to obtain the frequency optimization result. The frequency points of the initial detection signal evaluation and optimization model are defined using the frequency optimization results to obtain an amplitude optimization model. The amplitude optimization model is then solved to obtain the amplitude optimization results.

2. The two-stage optimization method for detection signals oriented towards load identification according to claim 1, characterized in that, The specific expression for the partial derivative matrix is: in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

3. The two-stage optimization method for detection signals oriented towards load identification according to claim 1, characterized in that, Solving the frequency-oriented optimization model to obtain the frequency optimization result includes: The decision variables of the frequency optimization model are randomly initialized within the range of their values ​​to obtain the first type group; Three individuals are randomly selected from the first population to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, a random number between 0 and 1 is generated. If the random number is less than the first preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target frequency function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population described above is executed iteratively until the target frequency function value converges, thus obtaining the frequency optimization result.

4. The two-stage optimization method for detection signals oriented towards load identification according to claim 1, characterized in that, Solving the magnitude-oriented optimization model to obtain the magnitude optimization result includes: The decision variables of the optimization model oriented towards magnitude optimization are randomly initialized within the range of their values ​​to obtain the second population; Three individuals are randomly selected from the second population to generate the next generation of individuals, where N is a positive integer; For each element of the current next generation individual, a random number between 0 and 1 is generated. If the random number is less than the second preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The crossover individuals are compared with their previous generation individuals, and the individuals with smaller target magnitude function values ​​are selected for the next generation of evolution; The generation, crossover, and selection processes are executed in parallel a preset number of times to obtain the next generation population; The process of generating the next generation population described above is executed iteratively until the target magnitude function value converges, thus obtaining the magnitude optimization result.

5. A two-stage optimization device for detection signals for load identification, characterized in that, include: An initial construction module is used to determine the set of discrete frequencies for frequency optimization, to calculate and store the partial derivative matrix, and to construct an initial probe signal evaluation and optimization model for load model identification based on the partial derivative matrix. The frequency optimization module is used to introduce a frequency penalty term into the initial detection signal evaluation and optimization model to obtain a frequency-optimized optimization model, and to solve the frequency-optimized optimization model to obtain the frequency optimization result. The amplitude optimization module is used to limit the frequency points of the initial detection signal evaluation and optimization model using the frequency optimization results, so as to obtain an optimization model for amplitude optimization, and to solve the optimization model for amplitude optimization to obtain the amplitude optimization result.

6. The two-stage optimization device for detection signals oriented towards load identification according to claim 5, characterized in that, The specific expression for the partial derivative matrix is: in, The partial derivative matrix, , Frequency The following parameters are The magnitude function and angle function of the transfer function of the load model. For the load model n One parameter, For the first i angular frequency, For the first i One frequency, The number of frequencies in the discrete frequency set.

7. The two-stage optimization device for detection signals oriented towards load identification according to claim 5, characterized in that, The frequency optimization module includes: The first initialization unit is used to randomly initialize the decision variables of the optimization model oriented towards frequency optimization within the range of their values, so as to obtain the first type group. The first generation unit is used to randomly select three individuals from the first population and iterate N times to generate the next generation of individuals, where N is a positive integer; The first crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the first preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The first evolutionary unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target frequency function value for the next generation of evolution; The first parallel unit is used to execute the generation, crossover, and selection processes a preset number of times in parallel to obtain the next generation population; The first iteration unit is used to iteratively execute the aforementioned process of generating the next generation population until the target frequency function value converges, thereby obtaining the frequency optimization result.

8. The two-stage optimization device for detection signals oriented towards load identification according to claim 5, characterized in that, The amplitude optimization module includes: The second initialization unit is used to randomly initialize the decision variables within the range of values ​​of the magnitude of the optimization model oriented towards magnitude optimization, so as to obtain the second population. The second generation unit is used to randomly select three individuals from the second population and iterate N times to generate the next generation of individuals, where N is a positive integer; The second crossover unit is used to generate a random number between 0 and 1 for each element of the current next generation individual. If the random number is less than the second preset crossover probability, crossover replacement is performed; otherwise, the original individual elements are retained, and the crossover individual is finally obtained. The second evolutionary unit is used to compare the crossover individual with its previous generation individual and select the individual with the smaller target amplitude function value for the next generation of evolution; The second parallel unit is used to execute the generation, crossover, and selection processes a preset number of times in parallel to obtain the next generation population; The second iteration unit is used to iteratively execute the aforementioned process of generating the next generation population until the target magnitude function value converges, thereby obtaining the magnitude optimization result.

9. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, the processor executing the program to implement the two-stage optimization method for load identification of detection signals as described in any one of claims 1-4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the two-stage optimization method for detection signals for load identification as described in any one of claims 1-4.

Images