Electrolytic aluminum frequency modulation method and device based on frequency response, medium and equipment

Abstract

The invention discloses an electrolytic aluminum frequency modulation method and device based on frequency response, a medium and equipment, and the method comprises the steps: firstly, dynamically updating a frequency response model parameter through the frequency actual measurement data after a system power fault, so as to calculate the electrolytic aluminum adjustment amount more accurately; and the current active power regulation quantity when electrolytic aluminum participates in frequency modulation is adjusted through iteration, so that the problem of frequency modulation resource shortage caused by output proportion change of different energy systems is effectively solved, the parameter change of the frequency response model can be considered in the regulation process, the regulation quantity calculation error is avoided, and the frequency of the power system is ensured to be effectively regulated and controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method, device, medium and equipment for frequency modulation of electrolytic aluminum based on frequency response. Background Art

[0002] Under the development trend of new power systems, the proportion of renewable energy electricity continues to rise, while thermal power units can no longer fully undertake the task of supporting system frequency due to the gradual decrease in their proportion and insufficient spare capacity. At the same time, high-energy-consuming industrial loads represented by electrolytic aluminum have the ability to participate in demand response. Such loads have the following characteristics: large rated power (usually up to hundreds of megawatts), wide power adjustment range and continuous adjustment, fast adjustment speed, and are heat storage loads, which have little impact on production when adjusting power. Therefore, electrolytic aluminum loads have become a highly potential frequency regulation resource.

[0003] However, most of the existing technologies focus on how to make high-energy loads such as electrolytic aluminum assist in primary frequency regulation. However, with the development of the power system, the proportion of thermal power units has gradually decreased, and frequency regulation resources have become increasingly scarce. At the same time, as the output ratio of different energy systems changes, the frequency response model parameters of the power system are also changing. At present, when applying the frequency response model to calculate the electrolytic aluminum regulation amount, the existing technology fails to fully consider the changes in these parameters, which may lead to errors in the calculation results of the electrolytic aluminum regulation amount. Summary of the invention

[0004] Based on this, it is necessary to provide an electrolytic aluminum frequency regulation method, device, medium and equipment based on frequency response to solve the problem that the proportion of thermal power units is gradually decreasing, the output share of different energy systems is changing, resulting in increasing scarcity of frequency regulation resources, and the failure to fully consider the changes in frequency response model parameters, resulting in errors in the calculation results of the electrolytic aluminum regulation amount.

[0005] A method for frequency modulation of electrolytic aluminum based on frequency response, the method comprising:

[0006] After a system power failure occurs, target parameters of the frequency response model are determined and imported based on the measured frequency data within a preset time period from the current time;

[0007] Based on the preset iteration step size, the active power regulation amount when the current electrolytic aluminum participates in the frequency modulation is adjusted for the kth time; wherein the initial value of the active power regulation amount is 0;

[0008] Substitute the current active power adjustment amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and judge whether the first adjustment condition and the second adjustment condition are met at present; wherein, the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than the preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold;

[0009] If at least one of the first adjustment condition and the second adjustment condition is not met, set k=k+1, and return to execute the step of adjusting the active power adjustment amount of the current electrolytic aluminum participating in frequency modulation for the kth time based on the preset iteration step and subsequent steps, until both the first adjustment condition and the second adjustment condition are met, and the electrolytic aluminum frequency is modulated based on the currently determined active power adjustment amount.

[0010] In one embodiment, after a system power failure occurs, determining and importing target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time includes:

[0011] Constructing an electric power simulation system on a simulation platform, and simulating the electric power simulation system under the premise of setting a frequency response model to different simulation parameters, so as to obtain frequency sample data corresponding to each simulation parameter;

[0012] After a system power failure occurs, the frequency measured data within a preset time period from the current time is obtained, and the distance between the frequency measured data and the frequency sample data corresponding to each simulation parameter is calculated, which is expressed as:

[0013]

[0014] In the above formula, M i represents the distance between the measured frequency data and the i-th frequency sample data, S(l) represents the l-th sampling value of the measured data S(l), L represents the total number of sampling values ​​of the measured data S(l), and Y i (l) represents the lth sampling value of the i-th frequency sample data;

[0015] The simulation parameters corresponding to the minimum distance are imported as the current target parameters.

[0016] In one embodiment, after a system power failure occurs, determining and importing target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time includes:

[0017] Constructing an electric power simulation system on a simulation platform, and simulating the electric power simulation system under the premise of setting a frequency response model to different initial simulation parameters, so as to obtain frequency sample data corresponding to each initial simulation parameter;

[0018] After a system power failure occurs, the actual frequency data within a preset time period from the current time is obtained, the initial error metric between the actual frequency data and the frequency sample data corresponding to each initial simulation parameter is calculated through the objective function, and the proxy model is fitted through Gaussian process regression; wherein the proxy model is used to indicate the functional relationship between each simulation parameter and the corresponding error metric;

[0019] Selecting multiple candidate simulation parameters from the proxy model, and using an acquisition function to calculate improved values ​​of the multiple candidate simulation parameters, setting the candidate simulation parameter with the largest improved value as the simulation parameter to be evaluated, and simulating the power simulation system under the premise of setting the frequency response model as the simulation parameter to be evaluated, so as to obtain frequency sample data corresponding to the simulation parameter to be evaluated; wherein the candidate simulation parameter is different from the initial simulation parameter;

[0020] The error metric to be evaluated between the frequency measured data and the frequency sample data corresponding to the simulation parameter to be evaluated is calculated through the objective function. If the error metric to be evaluated is greater than a preset error metric threshold, the proxy model is updated in combination with the error metric to be evaluated, and the step of selecting multiple candidate simulation parameters from the proxy model and subsequent steps are returned to execute until the error metric to be evaluated is less than or equal to the preset error metric threshold.

[0021] In one embodiment, the use of the acquisition function to calculate the improved values ​​of the plurality of candidate simulation parameters is expressed as:

[0022]

[0023] In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M best represents the currently known minimum error measure, ξ represents the exploration parameter, Φ(·) represents the cumulative distribution function of the standard normal distribution, and φ(·) represents the probability density function of the standard normal distribution.

[0024] In one embodiment, the use of the acquisition function to calculate the improved values ​​of the plurality of candidate simulation parameters is expressed as:

[0025]

[0026] In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M best represents the currently known minimum error measure, ξ represents the exploration parameter, and Φ(·) represents the cumulative distribution function of the standard normal distribution.

[0027] In one embodiment, the calculation formula of the transient frequency deviation in the frequency response model is as follows:

[0028]

[0029] In the above formula, Δω(t) represents the transient frequency deviation at time t, R represents the equivalent adjustment coefficient of the power system, ΔP represents the active power disturbance of the power system, D represents the load damping coefficient of the power system, and K m Represents the ratio of mechanical power to the installed capacity of the power system, ΔP AL represents the active power regulation, t0 represents the time when electrolytic aluminum participates in frequency modulation, α, ζ, ω n ,ω r and φ are the target parameters.

[0030] In one embodiment, the calculation formula of the frequency deviation extreme value in the frequency response model is:

[0031]

[0032] In the above formula, Δω max Indicates the frequency deviation extreme value, t z represents the time when the extreme value arrives, n represents the number of extreme value points, T R Represents the equivalent reheat time constant.

[0033] An electrolytic aluminum frequency modulation device based on frequency response, the electrolytic aluminum frequency modulation device based on frequency response comprising:

[0034] A model parameter determination module is used to determine and import target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time after a system power failure occurs;

[0035] An active power regulation amount determination module is used to adjust the active power regulation amount when the current electrolytic aluminum participates in frequency modulation for the kth time based on a preset iteration step; wherein the initial value of the active power regulation amount is 0;

[0036] A condition judgment module, used for substituting the current active power adjustment amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and judging whether the first adjustment condition and the second adjustment condition are currently met; if at least one of the first adjustment condition and the second adjustment condition is not met, then set k=k+1, and return to call the active power adjustment amount determination module; if both the first adjustment condition and the second adjustment condition are met, then call the frequency modulation module; wherein the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than a preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold;

[0037] The frequency modulation module is used to perform frequency modulation of electrolytic aluminum based on the currently determined active power adjustment amount.

[0038] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes the steps of the above-mentioned electrolytic aluminum frequency modulation method based on frequency response.

[0039] A terminal device includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above-mentioned electrolytic aluminum frequency modulation method based on frequency response.

[0040] The present invention provides a method, device, medium and equipment for electrolytic aluminum frequency modulation based on frequency response. First, the measured frequency data after the system power failure is used to dynamically update the frequency response model parameters so as to more accurately calculate the electrolytic aluminum adjustment amount. And by iteratively adjusting the active power adjustment amount when the current electrolytic aluminum participates in the frequency modulation, the problem of lack of frequency modulation resources caused by the change in the output proportion of different energy systems is effectively overcome, and the change of the frequency response model parameters can be considered in the adjustment process, avoiding the error in the calculation of the adjustment amount, and ensuring that the frequency of the power system is effectively regulated. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] in:

[0043] Figure 1 It is a schematic diagram of the frequency response process of the power system;

[0044] Figure 2It is a schematic diagram of the process of frequency modulation of electrolytic aluminum based on frequency response;

[0045] Figure 3a This is the electrolytic aluminum frequency modulation effect diagram with updated frequency model parameters when the proportion of new energy electricity is 0%;

[0046] Figure 3b This is the electrolytic aluminum frequency modulation effect diagram with updated frequency model parameters when the proportion of new energy electricity accounts for 25%;

[0047] Figure 3c This is the electrolytic aluminum frequency modulation effect diagram with updated frequency model parameters when the proportion of new energy electricity accounts for 50%;

[0048] Figure 4a This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity is 0%;

[0049] Figure 4b This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity accounts for 25%;

[0050] Figure 4c This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity accounts for 50%;

[0051] Figure 5 It is a structural schematic diagram of an electrolytic aluminum frequency modulation device based on frequency response;

[0052] Figure 6 This is a structural block diagram of the terminal equipment. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0055] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0056] The frequency response model of the power system was first proposed by PM Anderson and other scholars, who summarized the frequency response process of the synchronous unit power system as follows: Figure 1 As shown. Among them, P sp It represents the ratio of the impact load to the total load (when studying the frequency response after load disturbance, this value is set to 0), P m Represents mechanical power, P L represents the electromagnetic power of the generator, Δf represents the frequency increment of the power system, the above variables are all per unit values, and the benchmark capacity represents the total installed capacity of the power system S SB . F H It represents the capacity coefficient of high pressure cylinder of equivalent reheat unit, T R represents the equivalent reheating time constant, H represents the inertia time constant of the power system, D represents the load damping coefficient of the power system, including the damping coefficient of the generator and the frequency coefficient of the load, R represents the equivalent adjustment coefficient of the power system, K m It represents the ratio of mechanical power to the installed capacity of the power system and is a constant related to active reserve and power factor.

[0057] Under the development trend of new power systems, the proportion of renewable energy electricity continues to rise, while thermal power units can no longer fully undertake the task of supporting system frequency due to the gradual decrease in their proportion and insufficient spare capacity. At the same time, high-energy-consuming industrial loads represented by electrolytic aluminum have the ability to participate in demand response. Such loads have the following characteristics: large rated power (usually up to hundreds of megawatts), wide power adjustment range and continuous adjustment, fast adjustment speed, and are heat storage loads, which have little impact on production when adjusting power. Therefore, electrolytic aluminum loads have become a highly potential frequency regulation resource.

[0058] However, most of the existing technologies focus on how to make high-energy loads such as electrolytic aluminum assist in primary frequency regulation. However, with the development of the power system, the proportion of thermal power units has gradually decreased, and frequency regulation resources have become increasingly scarce. At the same time, as the output ratio of different energy systems changes, the frequency response model parameters of the power system are also changing. At present, when applying the frequency response model to calculate the electrolytic aluminum regulation amount, the existing technology fails to fully consider the changes in these parameters, which may lead to errors in the calculation results of the electrolytic aluminum regulation amount.

[0059] In response to the above problems, Figure 2 As shown, Figure 2 The present invention provides a schematic flow chart of an electrolytic aluminum frequency modulation method based on frequency response in an embodiment of the present invention. The steps provided by the electrolytic aluminum frequency modulation method based on frequency response in this embodiment include:

[0060] S201, after a system power failure occurs, target parameters of a frequency response model are determined and imported based on actual frequency measurement data within a preset time period from the current time.

[0061] Optionally, based on the existing technology, the frequency response model in this scenario is set as:

[0062]

[0063] Among them, α, ζ, ωn, ωr and φ are target parameters, that is, the parameters that need to be determined in this step, and the relationship they satisfy is as follows:

[0064]

[0065] The frequency response model describes the mathematical model of the frequency dynamic response process of the power system after a power failure (too large or too small). Through this model, it is possible to predict how frequency regulation resources such as electrolytic aluminum loads adjust system frequency under different circumstances.

[0066] Specifically, when a power failure occurs in the power system, the actual frequency data within a certain period of time is first obtained from the real-time monitoring system. Based on these data, the target parameters in the frequency response model are calculated by the least squares method or other fitting methods. Then, these target parameters are imported into the frequency response model for subsequent frequency prediction. In this way, the frequency response model parameters are dynamically adjusted according to the real-time frequency data, which can ensure that the power regulation amount of the electrolytic aluminum load accurately matches the system frequency demand and reduce the error in the regulation process.

[0067] S202, adjusting the active power regulation amount of the current electrolytic aluminum participating in frequency modulation for the kth time based on the preset iteration step size.

[0068] The preset iteration step Δx refers to the power adjustment amount increased or decreased each time the active power adjustment amount of the electrolytic aluminum load is adjusted. It is a preset value. Through the control of this iteration step, the system can fully adjust the adjustment power of the electrolytic aluminum. The initial value of the active power adjustment amount is 0. When the actual power demand of the power system exceeds the normal power supply capacity, the iteration step Δx is positive; when the actual power demand of the power system is less than the normal power supply capacity, the iteration step Δx is negative.

[0069] S203, substituting the current active power regulation amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment.

[0070] Among them, in the frequency response model, in order to make the electrolytic aluminum load participate in frequency modulation, the calculation formula of the transient frequency deviation before and after the electrolytic aluminum participates in regulation can be deduced from formula (1):

[0071]

[0072] In the above formula, Δω(t) represents the transient frequency deviation at time t, R represents the equivalent adjustment coefficient of the power system, ΔP represents the active power disturbance of the power system, D represents the load damping coefficient of the power system, and K m Represents the ratio of mechanical power to the installed capacity of the power system, ΔP AL represents the active power regulation, t0 represents the time when electrolytic aluminum participates in frequency modulation, α, ζ, ω n ,ω r and φ are the target parameters.

[0073] Furthermore, the formula for calculating the extreme value of frequency deviation in the frequency response model is deduced from formula (1):

[0074]

[0075]

[0076] In the above formula, Δω max Indicates the frequency deviation extreme value, t z represents the time when the extreme value arrives, n represents the number of extreme value points, T R Represents the equivalent reheat time constant.

[0077] S204: Determine whether the first adjustment condition and the second adjustment condition are currently met.

[0078] Among them, the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than the preset amplitude threshold. This condition is set to avoid excessive frequency fluctuations and ensure the stability of the frequency adjustment process.

[0079] The second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold. This condition is set to ensure that the system frequency is always kept within a safe range to avoid excessive frequency deviation from having an adverse effect on the power system. Optionally, the deviation threshold is set to 0.14.

[0080] If at least one of the first adjustment condition and the second adjustment condition is not satisfied, then S205 is executed, k=k+1 is set, and the process returns to S202 and subsequent steps. If both the first adjustment condition and the second adjustment condition are satisfied, then S206 is executed.

[0081] This step can ensure system stability while avoiding over-regulation and excessive frequency fluctuations, which effectively improves regulation efficiency and system responsiveness.

[0082] S206, performing frequency modulation of electrolytic aluminum based on the currently determined active power adjustment amount.

[0083] The system finally performs frequency regulation of the electrolytic aluminum load according to the active power regulation amount determined in the previous step. This step is the final implementation link in the frequency regulation process, ensuring that the frequency of the power system is effectively controlled to avoid excessive deviation, and that the electrolytic aluminum load can participate in the frequency regulation process.

[0084] The above-mentioned method for regulating the frequency of electrolytic aluminum based on frequency response first uses the measured frequency data after the system power failure to dynamically update the frequency response model parameters in order to more accurately calculate the regulation amount of electrolytic aluminum. And by iteratively adjusting the active power regulation amount when the current electrolytic aluminum participates in the frequency regulation, it effectively overcomes the problem of lack of frequency regulation resources caused by the change in the output proportion of different energy systems, and can consider the change of frequency response model parameters in the regulation process, avoiding the calculation error of the regulation amount, and ensuring that the frequency of the power system is effectively regulated.

[0085] Optionally, after a system power failure occurs, S201 determines target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time and imports the target parameters to specifically perform the following steps:

[0086] A1. Construct a power simulation system on a simulation platform, and simulate the power simulation system under the premise of setting the frequency response model to different simulation parameters to obtain frequency sample data corresponding to each simulation parameter.

[0087] Specifically, a complete power system model is established in a power simulation platform (such as RTDS) to simulate the power generation, load and regulation equipment of the system. Then, in the simulation system, the simulation parameters (such as grid load, generator type, regulation coefficient, etc.) are changed, and multiple simulations are run to generate frequency sample data corresponding to each simulation parameter.

[0088] A2. After a system power failure occurs, obtain the frequency measured data within a preset time from the current time, and calculate the distance between the frequency measured data and the frequency sample data corresponding to each simulation parameter, expressed as:

[0089]

[0090] In the above formula, M i represents the distance between the measured frequency data and the i-th frequency sample data, S(l) represents the l-th sampling value of the measured data S(l), L represents the total number of sampling values ​​of the measured data S(l), and Y i (l) represents the lth sampling value of the i-th frequency sample data.

[0091] A3. Import the simulation parameters corresponding to the minimum distance as the current target parameters.

[0092] It is understandable that as the proportion of new energy gradually increases, the frequency characteristics of the system will change. The use of this dynamic update method can enable the frequency response model to adapt to changes in system parameters, thereby ensuring the accuracy of frequency regulation resources such as electrolytic aluminum when participating in frequency regulation.

[0093] Optionally, to obtain more accurate target parameters, after a system power failure occurs, S201 determines the target parameters of the frequency response model according to the frequency measured data within a preset time period from the current time and imports the target parameters into the frequency response model. Specifically, the following steps are performed:

[0094] B1. Construct a power simulation system on a simulation platform, and simulate the power simulation system under the premise of setting the frequency response model to different initial simulation parameters to obtain frequency sample data corresponding to each initial simulation parameter.

[0095] It can be understood that since step B1 is basically the same as step A1 in another embodiment, it will not be described in detail here.

[0096] B2. After a system power failure occurs, obtain the actual frequency data within a preset time period from the current time, calculate the initial error measure between the actual frequency data and the frequency sample data corresponding to each initial simulation parameter through the objective function, and fit the proxy model through Gaussian process regression.

[0097] The objective function is a mathematical expression used to measure the difference between the model prediction and the measured data. For example, the error metric can be the mean absolute error, the maximum absolute error, etc., which can be set according to the actual situation. The proxy model is used to indicate the functional relationship between each simulation parameter and the corresponding error metric.

[0098]

[0099] Θ represents the simulation parameter, μ(θ) represents the prediction error measure at point θ, and σ(θ) represents the prediction uncertainty at point θ.

[0100] B3. Select multiple candidate simulation parameters from the proxy model, and use the acquisition function to calculate the improvement values ​​of the multiple candidate simulation parameters, set the candidate simulation parameter with the largest improvement value as the simulation parameter to be evaluated, and simulate the power simulation system under the premise of setting the frequency response model as the simulation parameter to be evaluated to obtain the frequency sample data corresponding to the simulation parameter to be evaluated.

[0101] Among them, multiple candidate simulation parameters are sampled from the proxy model at intervals or randomly. These candidate simulation parameters are different from the initial simulation parameters. The goal is to select additional parameters that are most likely to improve the error metric. The acquisition function is an important part of the optimization algorithm, which is used to determine the parameter point for the next optimization. It measures the degree of improvement (i.e., improvement value) of different candidate parameters relative to the current error metric, and selects the optimal candidate simulation parameter as the simulation parameter to be evaluated for evaluation.

[0102] Optionally, a collection function is used to calculate the improved values ​​of multiple candidate simulation parameters, expressed as:

[0103]

[0104] In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M best represents the currently known minimum error metric, ξ represents the exploration parameter, which is used to control the balance between exploration and exploitation. When ξ>0, there is more exploration, and when ξ=0, there is focus on exploitation. Φ(·) represents the cumulative distribution function of the standard normal distribution, and φ(·) represents the probability density function of the standard normal distribution. The goal of this acquisition function is to select a new evaluation point that maximizes its improvement expectation relative to the current optimal value, where μ(θ)-M best represents the potential improvement of the new point relative to the current optimal value, σ(θ) represents uncertainty, and the larger the standard deviation, the higher the uncertainty of the point, and the more likely it is to bring greater improvement. Φ(·) and φ(·) are used to measure the probability of improvement.

[0105] Optionally, a collection function is used to calculate the improved values ​​of multiple candidate simulation parameters, expressed as:

[0106]

[0107] In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M bestrepresents the currently known minimum error metric, ξ represents the exploration parameter, and Φ(·) represents the cumulative distribution function of the standard normal distribution. The goal of this acquisition function is to select a new evaluation point so that the probability that the objective function value at this point is greater than the current optimal value is maximized. The calculation is the standardized potential improvement size. Φ(·) represents the given current optimal value M best And the predicted mean μ(θ), the probability that the objective function value at this point is greater than the current optimal value.

[0108] B4. Calculate the error metric to be evaluated between the actual frequency data and the frequency sample data corresponding to the simulation parameter to be evaluated through the objective function. If the error metric to be evaluated is greater than the preset error metric threshold, update the proxy model in combination with the error metric to be evaluated, and return to execute the steps of B3 and subsequent steps until the error metric to be evaluated is less than or equal to the preset error metric threshold.

[0109] If the error metric to be evaluated is greater than the preset error metric threshold, then in each iteration, a new evaluation point is added to the set of all known simulation parameters, and these data are used to update the Gaussian process regression model. Specifically, the posterior distribution needs to be calculated and the hyperparameters of the surrogate model need to be updated.

[0110] The posterior distribution of Gaussian process regression is calculated using Bayes' theorem:

[0111]

[0112] Among them, μ(x) and Σ(x) represent the predicted mean and variance respectively, which can be calculated by the following formula:

[0113]

[0114] Among them, K is the matrix calculated by the covariance between the training data points, K * is the covariance vector between the new test point and the training data points, σ 2 is the variance of the noise and I is the identity matrix.

[0115] Through these steps, the Gaussian process regression model can be updated in combination with the new error metric to be evaluated, and then the process returns to execute step B3 and subsequent steps until the error metric to be evaluated is less than or equal to the preset error metric threshold.

[0116] In this way, the proxy model fitted by Gaussian process regression can accurately describe the relationship between simulation parameters and error metrics, thereby effectively selecting the optimal simulation parameters. This optimization method can continuously improve the accuracy of the frequency response model, so that the frequency response model can more accurately reflect the actual frequency changes of the system.

[0117] In order to verify the necessity of updating the frequency model parameters, this paper establishes an RTDS simulation model based on the actual power grid structure of a certain region, sets up system power failures under different proportions of renewable energy, and adopts the strategies of updating the frequency model parameters and not updating the frequency model parameters (selecting the average value of the parameters of all samples obtained by offline identification as the fixed value of the system model parameters without using the dynamic identification algorithm) to simulate the participation of electrolytic aluminum in power grid frequency control.

[0118] The simulation results of the electrolytic aluminum participating in the grid frequency control with the strategy of updating the frequency model parameters under different proportions of new energy are as follows: Figure 3a , Figure 3b , Figure 3c As shown. Among them, Figure 3a This is the electrolytic aluminum frequency modulation effect diagram with updated frequency model parameters when the proportion of new energy electricity is 0%; Figure 3b This is the electrolytic aluminum frequency modulation effect diagram with updated frequency model parameters when the proportion of new energy electricity accounts for 25%; Figure 3c This is the electrolytic aluminum frequency modulation effect diagram of updating the frequency model parameters when the proportion of new energy electricity is 50%; Figure 3a , Figure 3b , Figure 3c From the simulation results, it can be seen that under the premise of adopting the strategy of updating the frequency model parameters, even in the power system with a high proportion of renewable energy, the system transient frequency deviation can still be maintained within the deviation threshold (0.14Hz), and the adjustment amplitude of the transient frequency deviation at adjacent moments is small, thus realizing the precise control of the electrolytic aluminum load participating in the grid frequency regulation.

[0119] The simulation results of the control of the grid frequency by electrolytic aluminum with fixed frequency model parameters under different proportions of new energy are as follows: Figure 4a , Figure 4b , Figure 4c As shown. Among them, Figure 4a This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity is 0%; Figure 4b This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity accounts for 25%; Figure 4c This is the electrolytic aluminum frequency modulation effect diagram without updating the frequency model parameters when the proportion of new energy electricity accounts for 50%; Figure 4a , Figure 4b , Figure 4c The simulation results show that if the frequency model parameters are not updated in the calculation of the electrolytic aluminum regulation amount, even if the electrolytic aluminum frequency control is put into use, the frequency deviation of the transient process of the power system cannot be maintained within the deviation threshold (0.14Hz), and the adjustment range of the transient frequency deviation at adjacent moments is large, which cannot achieve accurate control of the electrolytic aluminum load participating in the grid frequency regulation.

[0120] Based on the above simulation results, it can be concluded that when electrolytic aluminum participates in grid frequency regulation, accurate control of the system frequency can only be achieved by identifying and updating the frequency response model parameters when a fault occurs.

[0121] In one embodiment, Figure 5 As shown, a frequency response-based electrolytic aluminum frequency modulation device is proposed, which includes:

[0122] A model parameter determination module 501 is used to determine and import target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time after a system power failure occurs;

[0123] The active power adjustment amount determination module 502 is used to adjust the active power adjustment amount when the current electrolytic aluminum participates in the frequency modulation for the kth time based on the preset iteration step; wherein the initial value of the active power adjustment amount is 0;

[0124] The condition judgment module 503 is used to substitute the current active power adjustment amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and judge whether the first adjustment condition and the second adjustment condition are currently met; if at least one of the first adjustment condition and the second adjustment condition is not met, set k=k+1, and return to call the active power adjustment amount determination module 502; if both the first adjustment condition and the second adjustment condition are met, call the frequency modulation module 504; wherein the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than a preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold;

[0125] The frequency modulation module 504 performs the frequency modulation of the electrolytic aluminum based on the currently determined active power adjustment amount.

[0126] Figure 6 FIG. 2 shows an internal structure diagram of a terminal device in an embodiment. Figure 6 As shown, the terminal device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the terminal device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement an electrolytic aluminum frequency modulation method based on frequency response. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may implement an electrolytic aluminum frequency modulation method based on frequency response. Those skilled in the art will understand that Figure 6The structure shown in the figure is only a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the terminal device to which the scheme of the present application is applied. The specific terminal device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0127] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented: after a system power failure occurs, the target parameters of a frequency response model are determined and imported according to the actual frequency measurement data within a preset time period from the current time; the active power adjustment amount when the current electrolytic aluminum participates in the frequency modulation is adjusted for the kth time based on the preset iteration step length; wherein the initial value of the active power adjustment amount is 0; the current active power adjustment amount is substituted into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and it is judged whether the first adjustment condition and the second adjustment condition are met at the current moment. Adjustment conditions; wherein, the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than the preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold; if at least one of the first adjustment condition and the second adjustment condition is not met, then set k=k+1, and return to execute the step and subsequent steps of adjusting the active power adjustment amount of the current electrolytic aluminum participating in the frequency modulation for the kth time based on the preset iteration step size, until the first adjustment condition and the second adjustment condition are both met, and the electrolytic aluminum frequency is modulated based on the currently determined active power adjustment amount.

[0128] A terminal device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: after a system power failure occurs, the target parameters of a frequency response model are determined and imported according to the actual frequency measurement data within a preset time period from the current time; the active power adjustment amount when the current electrolytic aluminum participates in the frequency modulation is adjusted for the kth time based on the preset iteration step length; wherein the initial value of the active power adjustment amount is 0; the current active power adjustment amount is substituted into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and it is judged whether the first modulation condition is satisfied at present. The first adjustment condition and the second adjustment condition; wherein, the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than the preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold; if at least one of the first adjustment condition and the second adjustment condition is not met, let k=k+1, and return to execute the step of adjusting the active power adjustment amount of the current electrolytic aluminum participating in the frequency modulation for the kth time based on the preset iteration step size and the subsequent steps, until the first adjustment condition and the second adjustment condition are both met, the electrolytic aluminum frequency modulation is performed based on the currently determined active power adjustment amount.

[0129] It should be noted that the above-mentioned method, device, equipment and computer-readable storage medium for electrolytic aluminum frequency modulation based on frequency response belong to a general inventive concept, and the contents in the embodiments of the method, device, equipment and computer-readable storage medium for electrolytic aluminum frequency modulation based on frequency response are applicable to each other.

[0130] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0131] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0132] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for frequency modulation of electrolytic aluminum based on frequency response, characterized in that: The method comprises: After a system power failure occurs, target parameters of the frequency response model are determined and imported based on the measured frequency data within a preset time period from the current time; Based on the preset iteration step size, the active power regulation amount when the current electrolytic aluminum participates in the frequency modulation is adjusted for the kth time; wherein the initial value of the active power regulation amount is 0; Substitute the current active power adjustment amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and judge whether the first adjustment condition and the second adjustment condition are met at present; wherein, the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than the preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold; If at least one of the first adjustment condition and the second adjustment condition is not met, set k=k+1, and return to execute the step of adjusting the active power adjustment amount of the current electrolytic aluminum participating in frequency modulation for the kth time based on the preset iteration step and subsequent steps, until both the first adjustment condition and the second adjustment condition are met, and the electrolytic aluminum frequency is modulated based on the currently determined active power adjustment amount.

2. The method according to claim 1, characterized in that After a system power failure occurs, the target parameters of the frequency response model are determined and imported according to the frequency measured data within a preset time period from the current time, including: Constructing an electric power simulation system on a simulation platform, and simulating the electric power simulation system under the premise of setting a frequency response model to different simulation parameters, so as to obtain frequency sample data corresponding to each simulation parameter; After a system power failure occurs, the frequency measured data within a preset time period from the current time is obtained, and the distance between the frequency measured data and the frequency sample data corresponding to each simulation parameter is calculated, which is expressed as: In the above formula, M i represents the distance between the measured frequency data and the i-th frequency sample data, S(l) represents the l-th sampling value of the measured data S(l), L represents the total number of sampling values ​​of the measured data S(l), and Y i (l) represents the lth sampling value of the i-th frequency sample data; The simulation parameters corresponding to the minimum distance are imported as the current target parameters.

3. The method according to claim 1, characterized in that After a system power failure occurs, the target parameters of the frequency response model are determined and imported according to the frequency measured data within a preset time period from the current time, including: Constructing an electric power simulation system on a simulation platform, and simulating the electric power simulation system under the premise of setting a frequency response model to different initial simulation parameters, so as to obtain frequency sample data corresponding to each initial simulation parameter; After a system power failure occurs, the actual frequency data within a preset time period from the current time is obtained, the initial error metric between the actual frequency data and the frequency sample data corresponding to each initial simulation parameter is calculated through the objective function, and the proxy model is fitted through Gaussian process regression; wherein the proxy model is used to indicate the functional relationship between each simulation parameter and the corresponding error metric; Selecting multiple candidate simulation parameters from the proxy model, and using an acquisition function to calculate improved values ​​of the multiple candidate simulation parameters, setting the candidate simulation parameter with the largest improved value as the simulation parameter to be evaluated, and simulating the power simulation system under the premise of setting the frequency response model as the simulation parameter to be evaluated, so as to obtain frequency sample data corresponding to the simulation parameter to be evaluated; wherein the candidate simulation parameter is different from the initial simulation parameter; The error metric to be evaluated between the frequency measured data and the frequency sample data corresponding to the simulation parameter to be evaluated is calculated through the objective function. If the error metric to be evaluated is greater than a preset error metric threshold, the proxy model is updated in combination with the error metric to be evaluated, and the step of selecting multiple candidate simulation parameters from the proxy model and subsequent steps are returned to execute until the error metric to be evaluated is less than or equal to the preset error metric threshold.

4. The method according to claim 3, characterized in that The use of the acquisition function to calculate the improved values ​​of the plurality of candidate simulation parameters is expressed as: In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M best represents the currently known minimum error measure, ξ represents the exploration parameter, Φ(·) represents the cumulative distribution function of the standard normal distribution, and φ(·) represents the probability density function of the standard normal distribution.

5. The method according to claim 3, characterized in that: The use of the acquisition function to calculate the improved values ​​of the plurality of candidate simulation parameters is expressed as: In the above formula, EI(θ) represents the improved value of simulation parameter θ, μ(θ) represents the prediction error measure at point θ, σ(θ) represents the prediction uncertainty at point θ, and M best represents the currently known minimum error measure, ξ represents the exploration parameter, and Φ(·) represents the cumulative distribution function of the standard normal distribution.

6. The method according to claim 1, characterized in that The calculation formula of the transient frequency deviation in the frequency response model is as follows, including: In the above formula, Δω(t) represents the transient frequency deviation at time t, R represents the equivalent adjustment coefficient of the power system, ΔP represents the active power disturbance of the power system, D represents the load damping coefficient of the power system, and K m Represents the ratio of mechanical power to the installed capacity of the power system, ΔP AL represents the active power regulation, t0 represents the time when electrolytic aluminum participates in frequency modulation, α, ζ, ω n ,ω r and φ are the target parameters.

7. The method according to claim 6, characterized in that The calculation formula of the frequency deviation extreme value in the frequency response model is: In the above formula, Δω max Indicates the frequency deviation extreme value, t z represents the time when the extreme value arrives, n represents the number of extreme value points, T R Represents the equivalent reheat time constant.

8. An electrolytic aluminum frequency modulation device based on frequency response, characterized in that: The electrolytic aluminum frequency modulation device based on frequency response comprises: A model parameter determination module is used to determine and import target parameters of a frequency response model based on actual frequency measurement data within a preset time period from the current time after a system power failure occurs; An active power regulation amount determination module is used to adjust the active power regulation amount when the current electrolytic aluminum participates in frequency modulation for the kth time based on a preset iteration step; wherein the initial value of the active power regulation amount is 0; A condition judgment module, used for substituting the current active power adjustment amount into the frequency response model to obtain the transient frequency deviation and the frequency deviation extreme value at the current moment, and judging whether the first adjustment condition and the second adjustment condition are currently met; if at least one of the first adjustment condition and the second adjustment condition is not met, then set k=k+1, and return to call the active power adjustment amount determination module; if both the first adjustment condition and the second adjustment condition are met, then call the frequency modulation module; wherein the first adjustment condition is that the adjustment amplitude between the transient frequency deviation at the current moment and the transient frequency deviation at the previous moment is less than a preset amplitude threshold, and the second adjustment condition is that the frequency deviation extreme value at the current moment is less than or equal to the preset deviation threshold; The frequency modulation module is used to perform frequency modulation of electrolytic aluminum based on the currently determined active power adjustment amount.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

10. A terminal device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 7.

Images