Electric power and electric quantity balance evaluation method and device based on new energy fluctuation

Abstract

The invention discloses a new energy fluctuation-based electric power and electric quantity balance evaluation method and device, and belongs to the technical field of electric power and electric quantity balance. When a singular value in performance parameter values of new energy is replaced in a performance parameter value mode, the number of parameters sent to the performance parameter value processing mode can affect the replacement of the singular value; and the accuracy of the obtained parameters is improved. If the number of the parameters sent to the performance parameter value processing mode is small, small singular values have a great effect on the accuracy of the parameters obtained by mode processing; if the number of parameters fed into the performance parameter value processing mode is large, the intensity of performing flexible processing on singular values is often not large, so that the feeding parameter number of the parameters processed by the most appropriate mode is obtained through the fluctuation amplitude of the parameters and the number of the singular values identified by the singular value algorithm. And the mode is used for adding the fed parameters, so that the obtained exported parameters are more accurate.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power and electricity balance, and specifically relates to an electric power and electricity balance evaluation method and device based on new energy fluctuations. Background Art

[0002] Power balance, also known as power system operation simulation, is used to study the improved operation mode of various power stations in the power system and the improved power exchange between subsystems, so as to verify the capacity and power efficiency of each plan. It is an important part of formulating power planning, design and operation plans.

[0003] In terms of the power and electricity balance of new energy fluctuations, it is often achieved by using the power and electricity balance method of new energy grid-connected operation with patent publication number "CN111987740B", which includes a generator set monitoring unit, a central control system and a power grid monitoring terminal that are sequentially communicated. The generator set monitoring unit is used to detect the performance parameter value of the new energy and detect whether the performance parameter value is within the normal range. When it is detected that the performance parameter value of the new energy is not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to promptly remind maintenance. The performance parameter value of the new energy is the power of the wind turbine (which can be the power per unit time), the temperature of the wind turbine, the wind power of the wind turbine, the power of the photovoltaic generator (which can be the power per unit time) or the temperature of the photovoltaic generator, so as to reduce the impact on the power and electricity balance.

[0004] The generator set monitoring unit samples the performance parameter values ​​of new energy through the corresponding detection module, and errors are often made in actual use. The current method of identifying the error values ​​(that is, singular values) in the performance parameter values ​​of new energy often identifies numerical points with large singular amplitudes in the performance parameter values ​​of new energy, and replaces them through the autoencoder, that is, through the singular value detection mode of the convolution layer and the pooling layer to process the performance parameter value queue arranged in the order of the sampling time points, and through corresponding learning, continuously improves the recognition function of the singular values ​​of the convolution layer and the pooling layer, and improves the singular value detection function of the model.

[0005] However, when the performance parameter values ​​of the new energy are sent to the autoencoder, even if the identified singular values ​​are relatively accurate, the performance parameter values ​​of the replaced new energy are often not accurate because the amount of singular values ​​in the performance parameter values ​​of the new energy sent is too large. Summary of the invention

[0006] In order to solve the defects existing in the prior art, the present invention proposes a method and device for evaluating the balance of electric power based on the fluctuation of new energy. When the singular values ​​in the performance parameter values ​​of new energy are replaced through the performance parameter value mode, the number of parameters sent to the performance parameter value processing mode will affect the accuracy of the parameters obtained after the singular values ​​are replaced. Just as, if the number of parameters sent to the performance parameter value processing mode is not large, then the few singular values ​​will have a significant effect on the accuracy of the parameters obtained by the mode processing; if the number of parameters sent to the performance parameter value processing mode is large, then the intensity of the flexible processing of the singular values ​​is often not large, so the number of parameters sent to the most appropriate mode to process the parameters is obtained through the fluctuation amplitude of the parameters themselves and the number of singular values ​​identified by the singular value algorithm, so that the mode performs addition on the sent parameters (that is, replaces the singular values ​​with the parameters formed by the mode), and the derived parameters obtained are more accurate.

[0007] The present invention utilizes the following technical solutions.

[0008] A method for evaluating power balance based on fluctuations of new energy sources, comprising:

[0009] The generator monitoring unit processes the performance parameter values ​​of the sampled new energy and detects whether the processed performance parameter values ​​are within the normal range. When it is detected that the performance parameter values ​​of the new energy are not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to remind maintenance in time. The performance parameter values ​​of the new energy are the power of the wind turbine, the temperature of the wind turbine, the wind power of the wind turbine, the power of the photovoltaic generator or the temperature of the photovoltaic generator, so as to reduce the impact on the balance of power and electricity;

[0010] The method for the generator set monitoring unit to process the performance parameter value of the sampled new energy source includes:

[0011] Step 1, sampling the performance parameter values ​​of new energy sources for a certain number of time intervals;

[0012] Step 2: For any time interval, determine the singular value probability of each numerical point, and perform grouping based on the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping;

[0013] Step 3: Calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm y The optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ;

[0014] Step 4, use a pre-defined performance parameter value processing mode to process the performance parameter value of the new energy to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

[0015] Furthermore, in Step 2, the numerical point is the performance parameter value of the new energy.

[0016] Furthermore, in Step 2, the algorithm for grouping numerical points is the K-means method, where the K value is six, and when grouping, the time point values ​​corresponding to all sampling moments in a time interval and the values ​​of the singular value probability of each numerical point are standardized using the Z-score method.

[0017] Furthermore, in Step 2, the method for determining the probability of singular values ​​of each numerical point is: obtain the number p of parameters whose k-th numerical point is a singular value at each time interval k Here, the piecewise linear regression method is used to determine whether the k-th numerical point is a singular value; the probability of the k-th numerical point being a singular value is obtained: Here, q k is the probability that the kth value point of each time interval is a singular value, and P is the number of time intervals;

[0018] Then, the method of determining whether a group is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping is: for any grouping, calculate the mean of the probabilities that each numerical point in the grouping is a singular value; if the mean is higher than the critical value, the grouping is determined to be a singular value grouping; if the mean is higher than the critical value, the grouping is determined not to be a singular value grouping.

[0019] Furthermore, in Step 2, the critical amount is thirty percent.

[0020] Furthermore, in Step 3, operation B y The equation is: v j is the number of parameters of all parameters in the jth singular value group within a time interval, N is the total number of singular value groups within a time interval, v U is the number of parameters defined in advance.

[0021] Furthermore, in Step 3, the operation C y The method is: calculate the standard deviation of the performance parameter value corresponding to the kth numerical point of all time intervals in the performance parameter value of the new energy The standard deviation of the k-th value point is standardized using the Z-score method to obtain the criticality factor b of the k-th value point at each time interval. k ; Get the self-information T of the kth numerical point of each time interval k; Faced with an arbitrary parameter subgroup, a number of operation parameter groups whose number of parameters is the number y of parameters fed into the model are sequentially extracted in the order of the time of the sampling value points within the time interval; Faced with an arbitrary operation parameter group, the cumulative value of the amount obtained by multiplying the criticality factor corresponding to each value point in the operation parameter group and the self-information amount is calculated to obtain the fluctuation amplitude of the operation parameter group;

[0022] Here, the self-information T of the kth numerical point at each time interval is k That is, take the kth value point as the midpoint, select the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u, and perform calculation on the kth value point and the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u to obtain the self-information T of the kth value point k .

[0023] Furthermore, in Step 4, the performance parameter value handling mode performs enhancement on the input parameters via Transformer so that the number of input parameters is the same as the number of output parameters.

[0024] Furthermore, Step 4 specifically includes:

[0025] Step 4-1, the construction number is the target input number y B The queue;

[0026] Step 4-2, facing any singular value group: if the number of queues is higher than the number of singular value groups, let the queues cover the singular value group to perform processing;

[0027] Step 4-3, facing any singular value group: if the number of queues is lower than the number of singular value groups, let the queue poll the singular value group. When polling for the first time, let the queue cover the local parameters of the head of the singular value group; during the polling period, move the queue, and the queue after the movement covers the local parameters of the singular value group that has not been polled.

[0028] An electric power balance assessment device based on new energy fluctuations, comprising:

[0029] The generator set monitoring unit, central control system and power grid monitoring terminal are connected in communication in sequence. The generator set monitoring unit is used to process the performance parameter value of the sampled new energy and detect whether the performance parameter value after processing is within the normal range. When it is detected that the performance parameter value of the new energy is not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to remind maintenance in time. The performance parameter value of the new energy is the power of the wind generator set, the temperature of the wind generator set, the wind force of the wind generator set, the power of the photovoltaic generator set or the temperature of the photovoltaic generator set, so as to reduce the impact on the balance of power and electricity;

[0030] The modules running on the generator set monitoring unit include:

[0031] A sampling module, which is used to sample the performance parameter value of the new energy for a certain period of time;

[0032] A grouping module is used to identify the singular value probability of each numerical point at any time interval, and perform grouping according to the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping according to the mean of the singular value probabilities of the numerical points in the grouping;

[0033] Operation module, which is used to calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm y The optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ;

[0034] The processing module is used to process the performance parameter value of the new energy by using a pre-defined performance parameter value processing mode to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

[0035] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0036] When replacing the singular values ​​in the performance parameter values ​​of new energy through the performance parameter value model, the number of parameters sent to the performance parameter value processing model will affect the accuracy of the parameters obtained after the singular values ​​are replaced. Just as, if the number of parameters sent to the performance parameter value processing model is small, then the small number of singular values ​​will have a significant impact on the accuracy of the parameters obtained by the model processing; if the number of parameters sent to the performance parameter value processing model is large, then the intensity of the flexible processing of the singular values ​​is often not large, so the number of input parameters of the parameters processed by the most appropriate model is obtained through the fluctuation amplitude of the parameters themselves and the number of singular values ​​identified by the singular value algorithm, so that the model adds to the input parameters (that is, replaces the singular values ​​with model-generated parameters), and the derived parameters obtained are more accurate.

[0037] By using the performance parameter values ​​of the new energy at each time interval, just as the performance parameter values ​​of the new energy sampled at a time interval are similar to the performance parameter values ​​of the new energy at another time interval, the singular values ​​will converge at the same sampling moment within a time interval. Just as the performance parameter values ​​of the sampled new energy have a high probability of producing singular values, several time intervals are sampled. The ratio of the probability that the sampled parameters at the same sampling moment in each time interval are singular values ​​is the probability that each time interval will produce an obstacle at the sampling moment. Therefore, by comparing the parameters of several time intervals, the singular value status of the set sampling moment can be more accurately identified, the problem of a single time interval parameter can be avoided, and the accuracy and reliability of the singular value probability can be improved.

[0038] The mean of all numerical points in a group is taken as the probability that the group is a singular value group. By comprehensively considering the singular value amplitudes of all groups, the singular value amplitudes of the group are more comprehensively determined. Then, by using the mean as the identification parameter, the effect of a single singular value numerical point on the identification of the group singular value can be reduced, thereby reducing the risk of misidentification and improving the accuracy of identification.

[0039] For singular value groups with a small number of parameters, the queue can directly cover them, so that there are regular parameters and singular values ​​here, and the singular values ​​are replaced through the performance parameter value processing mode. By allowing the queue to cover non-singular values ​​(regular parameters), it is ensured that the replaced parameters can better reflect the actual situation, making the parameters sent by the model more accurate. However, when facing singular value groups with a large number of parameters, the queue cannot cover them completely. At this time, the head part of the singular value group covered by the queue is polled. During the polling, the queue is moved, and the parameters in the queue are better maintained through local processing. The source attributes are maintained, thereby maintaining the completeness and continuity of the parameters.

[0040] The performance parameter value of the new energy of the present invention is a single category parameter and each value point has a sequential connection of sampling time, which is suitable for the corresponding singular value detection algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a partial flow chart of the electric power balance evaluation method based on the fluctuation of new energy sources described in the present invention;

[0042] Figure 2 It is a partial structural schematic diagram of the electric power balance assessment device based on new energy fluctuations described in the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the present invention clearer, the following will be combined with the drawings in the embodiments of the present invention to clearly and completely express the technical solution of the present invention. The embodiments expressed in this application are only some embodiments of the present invention, not all embodiments. According to the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the protection scope of the present invention.

[0044] like Figure 1 As shown, the electric power balance evaluation method based on the fluctuation of new energy sources described in the present invention comprises:

[0045] The generator set monitoring unit processes the performance parameter values ​​of the new energy obtained by sampling and detects whether the performance parameter values ​​after processing are within the normal range. When it is detected that the performance parameter values ​​of the new energy are not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to promptly remind maintenance. The performance parameter values ​​of the new energy are the power of the wind turbine set, the temperature of the wind turbine set, the wind force of the wind turbine set, the power of the photovoltaic generator set or the temperature of the photovoltaic generator set, so as to reduce the impact on the balance of power and electricity; the new energy is a photovoltaic generator set or a wind generator set.

[0046] The method for the generator set monitoring unit to process the performance parameter value of the sampled new energy source includes:

[0047] Step 1, sampling the performance parameter values ​​of new energy sources for a certain number of time intervals;

[0048] In a preferred but non-limiting embodiment of the present invention, in Step 1, the performance parameter value of the new energy is a single category parameter, and the performance parameter value of the new energy is the power of the wind turbine, the temperature of the wind turbine, the wind force of the wind turbine, the power of the photovoltaic generator or the temperature of the photovoltaic generator. The performance parameter value of the new energy corresponds to a performance parameter value, and the size of a time interval is 1 minute.

[0049] Step 2: For any time interval, determine the singular value probability of each numerical point, and perform grouping based on the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping;

[0050] In a preferred but non-limiting embodiment of the present invention, in Step 2, the numerical point is the performance parameter value of the new energy.

[0051] In a preferred but non-limiting embodiment of the present invention, in Step 2, the algorithm for grouping the numerical points is the K-means method, the K value of which is six, and the Z-score method is used to perform standardization on the time point values ​​corresponding to all sampling moments in a time interval and the values ​​of the singular value probability of each numerical point. And the numerical points in the group obtained by the K-means method continue in the order of the sampling moments, just as a group contains the numerical points sampled from the tenth to the twentieth in a time interval.

[0052] In a preferred but non-limiting embodiment of the present invention, in Step 2, the method for determining the probability of singular values ​​of each numerical point is: obtaining the number p of parameters whose k-th numerical point of each time interval is a singular value k Here, the piecewise linear regression method is used to determine whether the k-th numerical point is a singular value; the probability of the k-th numerical point being a singular value is obtained: Here, q k is the probability that the kth value point of each time interval is a singular value, and P is the number of time intervals;

[0053] Then, the method of determining whether a group is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping is: for any grouping, calculate the mean of the probabilities that each numerical point in the grouping is a singular value; if the mean is higher than the critical value, the grouping is determined to be a singular value grouping; if the mean is higher than the critical value, the grouping is determined not to be a singular value grouping.

[0054] In a preferred but non-limiting embodiment of the present invention, in Step 2, the critical amount is thirty percent.

[0055] Step 3: Calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm y The optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ;

[0056] Here, B y The number of performance parameter values ​​y sent to the new energy source and the number of parameters in each singular value group, C y The value of the numerical point is related to its self-information, e is the Euler number, and the reasonable amplitude factor Q y and B y Proportional to the reasonable amplitude factor Q y and C y Directly proportional.

[0057] In a preferred but non-limiting embodiment of the present invention, in Step 3, operation B y The equation is: v jis the number of parameters of all parameters in the jth singular value group within a time interval, N is the total number of singular value groups within a time interval, v U is the number of predefined parameters. Here, B y It is proportional to the square of the number of parameters of all parameters in each singular value group. y The lower it is, the smaller the ratio of parameters in the singular value group within the input parameters is, and the greater the accuracy of the derived parameters obtained by adding the input parameters is.

[0058] In a preferred but non-limiting embodiment of the present invention, in Step 3, the operation C y The method is: calculate the standard deviation of the performance parameter value corresponding to the kth numerical point of all time intervals in the performance parameter value of the new energy The standard deviation of the k-th value point is standardized using the Z-score method to obtain the criticality factor b of the k-th value point at each time interval. k ; Get the self-information T of the kth numerical point of each time interval k ; Faced with an arbitrary parameter subgroup, a number of operation parameter groups whose number of parameters is the number y of parameters fed into the model are sequentially extracted in the order of the time of the sampling value points within the time interval; Faced with an arbitrary operation parameter group, the cumulative value of the amount obtained by multiplying the criticality factor corresponding to each value point in the operation parameter group and the amount of self-information is calculated to obtain the fluctuation amplitude of the operation parameter group; The maximum amount of the fluctuation amplitude in the entire operation parameter group is C y Here, the lower Bx is, the lower the fluctuation of the parameters sent to the model within the parameters sampled in a time interval (the standard deviation corresponding to the numerical point multiplied by the amount of self-information is not large), and the accuracy of the derived parameters sent by the model is higher. The parameters sent by the model are the parameters sent by the performance parameter value processing model, and the derived parameters sent by the model are the parameters sent by the performance parameter value processing model;

[0059] Here, the self-information T of the kth numerical point at each time interval is k That is, take the kth value point as the midpoint, select the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u, and perform calculation on the kth value point and the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u to obtain the self-information T of the kth value point k The method for obtaining the amount of self-information is an existing method and will not be described in detail here.

[0060] In short, in B y The lower the Q y The lower the value, the greater the accuracy of the parameters derived from the model; y The lower the Q yThe lower the value, the greater the accuracy of the parameters derived from the model. y The higher the Q y The higher the value, the less accurate the parameters derived from the model; y The higher the Q y The higher the value, the less accurate the parameters derived from the model.

[0061] Step 4, use a pre-defined performance parameter value processing mode to process the performance parameter value of the new energy to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

[0062] In a preferred but non-limiting embodiment of the present invention, in Step 4, the performance parameter value handling mode performs enhancement on the input parameters via the Transformer so that the number of input parameters is the same as the number of output parameters.

[0063] Here, the number of parameters sent into the performance parameter value handling mode is the target number y B , the singular value is the numerical point within the performance parameter value of the new energy that belongs to the singular value grouping.

[0064] In a preferred but non-limiting embodiment of the present invention, Step 4 specifically includes:

[0065] Step 4-1, the construction number is the target input number y B The queue;

[0066] Step 4-2, facing any singular value group: if the number of queues is higher than the number of singular value groups, let the queues cover the singular value group to perform processing;

[0067] Step 4-3, facing any singular value group: if the number of queues is lower than the number of singular value groups, let the queue poll the singular value group. When polling for the first time, let the queue cover the local parameters of the head of the singular value group; during the polling period, move the queue, and the queue after the movement covers the local parameters of the singular value group that has not been polled.

[0068] Specifically, it is necessary to replace the numerical points belonging to the singular value group in a time interval (remove them and form new data through the numerical points of the non-singular value group). When the number of parameters in a singular value group is lower than the number of queues, align the midpoint of the queue with the midpoint of the parameters in the singular value group, and send the parameters in the queue as the input parameters into the model. When the number of parameters in a singular value group is higher than the number of queues, let the queue cover the local parameters of the head of the singular value group, and send the parameters in the queue as the input parameters into the model (the model can be an autoencoder), then replace the derived parameters of the model with the source parameters in the queue, and move the queue to let the queue cover the parameters belonging to the singular value group that have not been processed, until the queue polls all singular value groups.

[0069] Here, the number of queue parts covering the first segment of the singular value grouping is not higher than half of the total number of queues, so that the parameter ratio in the singular value grouping is small and the parameters derived from the model are more accurate.

[0070] like Figure 2 As shown, the electric power balance assessment device based on the fluctuation of new energy sources described in the present invention comprises:

[0071] The generator set monitoring unit, the central control system and the power grid monitoring terminal are sequentially connected in communication. The generator set monitoring unit is used to process the performance parameter value of the sampled new energy and detect whether the performance parameter value after processing is within the normal range. When it is detected that the performance parameter value of the new energy is not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to remind maintenance in time. The performance parameter value of the new energy is the power of the wind generator set (which can be the power per unit time), the temperature of the wind generator set, the wind power of the wind generator set, the power of the photovoltaic generator set (which can be the power per unit time) or the temperature of the photovoltaic generator set, so as to reduce the impact on the balance of power and electricity;

[0072] The modules running on the generator set monitoring unit include:

[0073] A sampling module, which is used to sample the performance parameter value of the new energy for a certain period of time;

[0074] A grouping module is used to identify the singular value probability of each numerical point at any time interval, and perform grouping according to the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping according to the mean of the singular value probabilities of the numerical points in the grouping;

[0075] Operation module, which is used to calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm yThe optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ;

[0076] The processing module is used to process the performance parameter value of the new energy by using a pre-defined performance parameter value processing mode to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

[0077] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0078] When replacing the singular values ​​in the performance parameter values ​​of new energy through the performance parameter value model, the number of parameters sent to the performance parameter value processing model will affect the accuracy of the parameters obtained after the singular values ​​are replaced. Just as, if the number of parameters sent to the performance parameter value processing model is small, then the small number of singular values ​​will have a significant impact on the accuracy of the parameters obtained by the model processing; if the number of parameters sent to the performance parameter value processing model is large, then the intensity of the flexible processing of the singular values ​​is often not large, so the number of input parameters of the parameters processed by the most appropriate model is obtained through the fluctuation amplitude of the parameters themselves and the number of singular values ​​identified by the singular value algorithm, so that the model adds to the input parameters (that is, replaces the singular values ​​with model-generated parameters), and the derived parameters obtained are more accurate.

[0079] By using the performance parameter values ​​of the new energy at each time interval, just as the performance parameter values ​​of the new energy sampled at a time interval are similar to the performance parameter values ​​of the new energy at another time interval, the singular values ​​will converge at the same sampling moment within a time interval. Just as the performance parameter values ​​of the sampled new energy have a high probability of producing singular values, several time intervals are sampled. The ratio of the probability that the sampled parameters at the same sampling moment in each time interval are singular values ​​is the probability that each time interval will produce an obstacle at the sampling moment. Therefore, by comparing the parameters of several time intervals, the singular value status of the set sampling moment can be more accurately identified, the problem of a single time interval parameter can be avoided, and the accuracy and reliability of the singular value probability can be improved.

[0080] The mean of all numerical points in a group is taken as the probability that the group is a singular value group. By comprehensively considering the singular value amplitudes of all groups, the singular value amplitudes of the group are more comprehensively determined. Then, by using the mean as the identification parameter, the effect of a single singular value numerical point on the identification of the group singular value can be reduced, thereby reducing the risk of misidentification and improving the accuracy of identification.

[0081] For singular value groups with a small number of parameters, the queue can directly cover them, so that there are regular parameters and singular values ​​here, and the singular values ​​are replaced through the performance parameter value processing mode. By allowing the queue to cover non-singular values ​​(regular parameters), it is ensured that the replaced parameters can better reflect the actual situation, making the parameters sent by the model more accurate. However, when facing singular value groups with a large number of parameters, the queue cannot cover them completely. At this time, the head part of the singular value group covered by the queue is polled. During the polling, the queue is moved, and the parameters in the queue are better maintained through local processing. The source attributes are maintained, thereby maintaining the completeness and continuity of the parameters.

[0082] The performance parameter value of the new energy of the present invention is a single category parameter and each value point has a sequential connection of sampling time, which is suitable for the corresponding singular value detection algorithm.

[0083] The present disclosure can be a system, method and / or computer program product. The computer program product can include a computer-readable backup medium carrying computer-readable program instructions for causing a processor to implement various aspects of the disclosure.

[0084] The computer readable backup medium can be a tangible network capable of retaining and backing up the instructions used by the instruction execution network. The computer readable backup medium can be, but is not limited to, an electrical backup network, a magnetic backup network, an optical backup network, an electromagnetic backup network, a semiconductor backup network, or any suitable combination of the above. Further examples of computer readable backup media (non-exhaustive list) include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disk read-only memory (HD-ROM), digital versatile disk (DXD), memory stick, floppy disk, mechanically encoded network, such as punch cards or protrusions in grooves with instructions backed up thereon, and any suitable combination of the above. The computer readable backup medium used herein is not to be construed as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a power line cable), or an electrical signal transmitted through wires.

[0085] The computer readable program instructions expressed herein can be downloaded from the computer readable backup medium to each inference / processing grid line, or downloaded to an external computer or external backup grid line through a wireless network, such as the Internet, a local area network, a wide area network and / or a wireless network. The wireless network can include copper transmission cables, power transmission lines, wireless transmission, routers, firewalls, switches, gateway computers and / or edge service devices. The wireless network adapter card or wireless network interface in each inference / processing grid line receives the computer readable program instructions from the wireless network and forwards the computer readable program instructions for storage in the computer readable backup medium in each inference / processing grid line.

[0086] The computer program instructions for performing the operations of the present disclosure can be assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, conditional setting values, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Sdaletal A, H++, etc., and conventional procedural programming languages ​​such as "H" language or similar programming languages. The computer-readable program instructions can be executed entirely on the client computer, partially on the client computer, as a separate software package, partially on the client computer and partially on the remote computer, or completely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the client computer through any other wireless network, including a local area network (LAb) or a wide area network (WAb), or can be connected to an external computer (just like using an Internet service provider to connect through the Internet). In some embodiments, the electronic circuit is customized by using the status values ​​of computer-readable program instructions, such as a programmable logic circuit, a field programmable gate array (FPGA) or a programmable logic array (PLA), and the electronic circuit can execute computer-readable program instructions to achieve various aspects of the overhead disclosure.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not deviate from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for evaluating power balance based on fluctuations in new energy, characterized in that: include: The generator monitoring unit processes the performance parameter values ​​of the sampled new energy and detects whether the processed performance parameter values ​​are within the normal range. When it is detected that the performance parameter values ​​of the new energy are not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to remind maintenance in time. The performance parameter values ​​of the new energy are the power of the wind turbine, the temperature of the wind turbine, the wind power of the wind turbine, the power of the photovoltaic generator or the temperature of the photovoltaic generator, so as to reduce the impact on the balance of power and electricity; The method for the generator set monitoring unit to process the performance parameter value of the sampled new energy source includes: Step 1, sampling the performance parameter values ​​of new energy sources for a certain number of time intervals; Step 2: For any time interval, determine the singular value probability of each numerical point, and perform grouping based on the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping; Step 3: Calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm y The optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ; Step 4, use a pre-defined performance parameter value processing mode to process the performance parameter value of the new energy to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

2. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 1 is characterized in that: In Step 2, the numerical point is the performance parameter value of the new energy.

3. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 2 is characterized in that: In Step 2, the algorithm for grouping numerical points is the K-means method, where the K value is six, and when grouping, the time point values ​​corresponding to all sampling moments in a time interval and the values ​​of the singular value probability of each numerical point are standardized using the Z-score method.

4. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 3 is characterized in that: In Step 2, the method to determine the probability of singular values ​​of each numerical point is: obtain the number p of parameters whose k-th numerical point is a singular value at each time interval k Here, the piecewise linear regression method is used to determine whether the k-th numerical point is a singular value; the probability of the k-th numerical point being a singular value is obtained: Here, q k is the probability that the kth value point of each time interval is a singular value, and P is the number of time intervals; Then, the method of determining whether a group is a singular value grouping based on the mean of the singular value probabilities of the numerical points in the grouping is: for any grouping, calculate the mean of the probabilities that each numerical point in the grouping is a singular value; if the mean is higher than the critical value, the grouping is determined to be a singular value grouping; if the mean is higher than the critical value, the grouping is determined not to be a singular value grouping.

5. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 4 is characterized in that: In Step 2, the critical amount is thirty percent.

6. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 5 is characterized in that: In Step 3, operation B y The equation is: v j is the number of parameters of all parameters in the jth singular value group within a time interval, N is the total number of singular value groups within a time interval, v U is the number of parameters defined in advance.

7. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 6 is characterized in that: In Step 3, calculate C y The method is: calculate the standard deviation of the performance parameter value corresponding to the kth numerical point of all time intervals in the performance parameter value of the new energy The standard deviation of the k-th value point is standardized using the Z-score method to obtain the criticality factor b of the k-th value point at each time interval. k ; Get the self-information T of the kth numerical point of each time interval k ; Faced with an arbitrary parameter subgroup, a number of operation parameter groups whose number of parameters is the number y of parameters fed into the model are sequentially extracted in the order of the time of the sampling value points within the time interval; Faced with an arbitrary operation parameter group, the cumulative value of the amount obtained by multiplying the criticality factor corresponding to each value point in the operation parameter group and the self-information amount is calculated to obtain the fluctuation amplitude of the operation parameter group; Here, the self-information T of the kth numerical point at each time interval is k That is, take the kth value point as the midpoint, select the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u, and perform calculation on the kth value point and the performance parameter value of the new energy source whose number of adjacent sampling moments on both sides is u to obtain the self-information T of the kth value point k .

8. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 7 is characterized in that: In Step 4, the performance parameter value handling model performs enhancement on the input parameters through Transformer so that the number of input parameters is the same as the number of output parameters.

9. The method for evaluating the power balance based on the fluctuation of new energy sources according to claim 8 is characterized in that: Step 4 specifically includes: Step 4-1, the construction number is the target input number y B The queue; Step 4-2, facing any singular value group: if the number of queues is higher than the number of singular value groups, let the queues cover the singular value group to perform processing; Step 4-3, facing any singular value group: if the number of queues is lower than the number of singular value groups, let the queue poll the singular value group. When polling for the first time, let the queue cover the local parameters of the head of the singular value group; during the polling period, move the queue, and the queue after the movement covers the local parameters of the singular value group that has not been polled.

10. A power balance assessment device based on new energy fluctuations, characterized in that: include: The generator set monitoring unit, central control system and power grid monitoring terminal are connected in communication in sequence. The generator set monitoring unit is used to process the performance parameter value of the sampled new energy and detect whether the performance parameter value after processing is within the normal range. When it is detected that the performance parameter value of the new energy is not within the normal range, the detection information is transmitted to the power grid monitoring terminal via the central control system to remind maintenance in time. The performance parameter value of the new energy is the power of the wind generator set, the temperature of the wind generator set, the wind force of the wind generator set, the power of the photovoltaic generator set or the temperature of the photovoltaic generator set, so as to reduce the impact on the balance of power and electricity; The modules running on the generator set monitoring unit include: A sampling module, which is used to sample the performance parameter value of the new energy for a certain period of time; A grouping module is used to identify the singular value probability of each numerical point at any time interval, and perform grouping according to the singular value probability to obtain a pre-defined number of groups; for any grouping, determine whether the grouping is a singular value grouping according to the mean of the singular value probabilities of the numerical points in the grouping; Operation module, which is used to calculate the reasonable amplitude factor Q y : Calculate the reasonable amplitude factor Q through the greedy algorithm y The optimal solution is used to identify the corresponding number of feeds y, which is defined as the target number of feeds y B ; The processing module is used to process the performance parameter value of the new energy by using a pre-defined performance parameter value processing mode to obtain the performance parameter value of the new energy after replacing the singular value, and use it as the processed performance parameter value.

Images