A user-side heat-electricity load prediction method and a terminal device

Abstract

The application is suitable for the field of load prediction, and provides a user-side heat and power load prediction method and terminal equipment. The method comprises the following steps: acquiring heat and power load data; based on the heat and power load data, a multi-layer partitioned space load model is constructed under the condition of distributed energy constraints and heat and power balance constraints; the load data of each layer in the multi-layer partitioned space load model is divided into total load data, classified load data, middle area load data and small area load data; the multi-layer partitioned space load model is initialized; based on the initialized data of the multi-layer partitioned space load model, the multi-layer partitioned space load model is iteratively solved; and if the iterative solving result meets preset conditions, a space load prediction value is output. The user-side heat and power load prediction method can improve the accuracy of load distribution prediction results.

Description

Technical Field

[0001] This application belongs to the field of load forecasting, and in particular relates to a user-side thermal and electrical load forecasting method and terminal equipment. Background Technology

[0002] In practical planning work, planners need to predict not only the total load but also the specific growth and distribution of the load. This is the main research content of spatial load forecasting. Economic, climatic, environmental, and many other uncertainties affect the load. In particular, due to the popularization of new energy sources and the construction of integrated energy systems, a large number of distributed power sources and thermoelectric coupling devices have been added to the power system. The uncertainties of distributed power sources and the thermoelectric balance constitute a complex, nonlinear mapping relationship with the load curve. Therefore, load forecasting not only needs to consider electrical load but also, by taking into account various factors and accurately predicting the integrated energy load, has become the biggest challenge in load forecasting research.

[0003] With the large-scale adoption of information technology in the power system and various sectors of the national economy, collecting historical load data and related factors has become much easier than before. Coupled with continuously improving intelligent computing methods, this provides valuable tools and research methods for comprehensive energy load forecasting. Therefore, based on the massive historical data of user-side thermal and power loads and distributed energy resources, it is possible to deeply analyze the dynamic coupling degree between distributed energy resources and thermal and power, establish a comprehensive energy load characteristic analysis model, and provide a reliable basis for power grid planning and operation.

[0004] However, the data on distributed energy resources, user-side thermoelectric loads, and related influencing factors are large in volume, lack regularity, and are difficult to classify. Furthermore, distributed power sources exhibit significant randomness and vary considerably across different regions, and the coupling characteristics of thermoelectric loads also differ significantly from region to region. Based on spatial analysis of the characteristics of distributed power sources and thermoelectric coupled loads, a key challenge in this research is how to classify historical data on distributed power sources and thermoelectric loads at different levels and across different regions.

[0005] Traditional spatial load forecasting models typically employ a two-layer partitioning structure, meaning they only have data on total load and local load. Relying solely on the mutual correction between local and total loads makes it difficult to obtain accurate load distribution forecasts, and land partitioning alone cannot account for the impact of distributed power generation and thermoelectric loads. Summary of the Invention

[0006] To overcome the problems existing in related technologies, this application provides a user-side thermoelectric load prediction method and terminal equipment, which can improve the accuracy of load distribution prediction results.

[0007] This application is achieved through the following technical solution:

[0008] In a first aspect, embodiments of this application provide a user-side thermoelectric load prediction method, including:

[0009] Acquire thermal load data;

[0010] Based on the aforementioned thermal and electrical load data, a multi-layered zonal spatial load model is constructed under distributed energy constraints and thermal and electrical balance constraints. The load data of each layer in the multi-layered zonal spatial load model is divided into total load data, classified load data, central zone load data, and small zone load data.

[0011] Initialize the multi-layered partitioned spatial load model;

[0012] Based on the initial data of the multi-level partitioned spatial load model, the multi-level partitioned spatial load model is solved iteratively.

[0013] If the iterative solution result meets the preset conditions, the spatial load prediction value is output.

[0014] In one possible implementation of the first aspect, acquiring the thermoelectric load data includes:

[0015] The acquisition of thermoelectric load data includes:

[0016] Obtain raw data on thermal and electrical loads;

[0017] Abnormal data points in the original thermoelectric load data are identified and corrected by longitudinal comparison and lateral comparison methods to obtain the thermoelectric load data.

[0018] In one possible implementation of the first aspect, the construction of a multi-layered zonal spatial load model under distributed energy condition constraints and thermoelectric balance constraints includes:

[0019] The heat and power load data is configured with load intervals, and consistency processing is performed on the load intervals; the load intervals are divided into total load intervals, category load intervals, mid-zone load intervals, and sub-zone load intervals.

[0020] Determine the constraints of distributed power source conditions and thermoelectric coupling load conditions;

[0021] A multi-layered spatial load model is constructed based on the load range, the distributed power source constraints, and the thermoelectric coupling load constraints.

[0022] In one possible implementation of the first aspect, setting the load range for the processed raw thermoelectric load data includes:

[0023] Let t∈{1,2,…,T} be the load classification number, and the data for each layer include: total load interval [L - s,L + s ], the total load interval of type t [L - st ,L + st ], Central load interval [L - ma ,L + ma ] and the load range of the community [L - sa ,L + sa ];

[0024] The consistency processing of the load range includes:

[0025] pass

[0026]

[0027] Make the classification load L st The aggregated value should be within the total load range [L] - s ,L + s [Inside]; where Δ(.,.) represents the calculation of the total load after the classified loads are weighted and superimposed according to the load curves; L st C represents the classification load of class t; st =[C st (0) C st (1) ,…,C st (i) ,…,C st (288) ] represents the 288-point load curve sampled every five minutes on a typical day for the t-th category of load, where C st (i) ∈[0,100], C st (i) It is the ratio after dividing by the largest element in the vector as the denominator;

[0028] pass

[0029]

[0030] The total load value of all the aforementioned central zones should be within the total load range; wherein, L ma (m) represents the central zone load numbered m; α ma-s The simultaneity rate from the central load to the total load;

[0031] pass

[0032]

[0033] The summation value of the load of any of the aforementioned cells within the central area is within the load interval of the central area; L - ma ,L + ma Indicates the upper and lower limits of the load in the middle zone; L sa For any cell within the central area;

[0034] pass

[0035]

[0036]

[0037] Ensure that the aggregated load value of the cells in class t falls within the total load range of that class; α sa-s,t The concurrency rate when summarizing the load of cell type t;

[0038] The consistency processing can obtain the optimal value for each load interval.

[0039] In one possible implementation of the first aspect, the distributed energy condition constraints include: wind power reactive power constraints and energy storage system operation constraints.

[0040] The reactive power constraint for wind power generation is:

[0041]

[0042] Among them, P w The active power output of the wind turbine is tanδ, the active power coefficient of the wind turbine is X. σ X is the sum of the generator stator reactance and rotor reactance. m R is the magnetizing reactance, R is the rotor resistance, and s is the slip.

[0043] The operating constraints of the energy storage system are:

[0044]

[0045]

[0046] in, and Let t be the maximum and minimum heat storage capacity of the thermal storage system. For different times, the heating temperature and recuperation temperature at the nodes of the energy storage system will change, and the SOC range of the energy storage system will also change to some extent. and These are the maximum and minimum energy storage values ​​of the electrical energy storage system.

[0047] In one possible implementation of the first aspect, the thermoelectric coupling load condition constraint is:

[0048]

[0049]

[0050] Where, Φ t It is the sum of the system heat load and network heat loss of the thermal system in time period t, obtained through power flow calculation; N h For all centralized heat sources and CHP unit nodes of the system; N e N represents the set of all generator nodes in the system. b P represents the set of all distributed heat pump and electric boiler nodes in the system. t It is the sum of the system power load and network loss in the power system during time period t, which is obtained through power flow calculation; These are the thermal energy and electrical energy emitted by the energy source at node j, respectively, and x j Its running status; and These represent the heat absorption and charging amount of the thermal and electrical energy storage devices at time t, respectively. and η represents the heat release and discharge of the thermal and electrical energy storage devices during time period t, respectively. qdis and η pdis These are the energy release efficiencies of thermal energy storage devices and electrical energy storage devices, respectively.

[0051] In one possible implementation of the first aspect, initializing the multi-layered partitioned spatial load model includes:

[0052] Verify the multi-layered spatial load model;

[0053] The iterative error initialization is performed on the verified multi-level partitioned spatial load model; the iterative error initialization includes: iterative forgetting weight matrix initialization, covariance matrix initialization, and partitioned load weight initialization; the forgetting weight matrix, the covariance matrix, and the partitioned load weight are the parameters of the iterative error.

[0054] In one possible implementation of the first aspect, the iterative solution of the multi-level partitioned spatial load model based on the initialization data of the multi-level partitioned spatial load model includes:

[0055] Based on the iterative error initialization data of the multi-level partitioned spatial load model, the residual of the multi-level partitioned spatial load model is obtained;

[0056] Based on the residuals of the multi-level partitioned spatial load model, the objective function of the multi-level partitioned spatial load model is obtained.

[0057] Minimize the objective function;

[0058] Minimizing the objective function includes:

[0059] Update the iteration error and update the cell load;

[0060] Based on the updated iteration error and cell load, the residuals of the multi-level partitioned spatial load model are updated.

[0061] In one possible implementation of the first aspect, the preset condition is:

[0062] max(|ε(t)-ε(t-1)|)<E max

[0063] Among them, E max The threshold is set to 0.001 in the example, and t represents the number of iterations.

[0064] Secondly, embodiments of this application provide a user-side thermoelectric load prediction device, including: an acquisition module for acquiring thermoelectric load data;

[0065] The construction module is used to construct a multi-layered zonal spatial load model based on the heat and power load data, under the constraints of distributed energy conditions and heat and power balance; the load data of each layer in the multi-layered zonal spatial load model is divided into total load data, classified load data, central zone load data and small zone load data;

[0066] An initialization module is used to initialize the multi-layer partitioned spatial load model;

[0067] The solution module is used to iteratively solve the multi-level partitioned spatial load model based on the initialization data of the multi-level partitioned spatial load model.

[0068] The output module is used to output the spatial load prediction value if the iterative solution result meets the preset conditions.

[0069] Thirdly, embodiments of this application provide a terminal device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, characterized in that the processor executes the computer program to implement the user-side thermoelectric load prediction method as described in any of the first aspects.

[0070] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, characterized in that, when executed by a processor, the computer program implements the user-side thermoelectric load prediction method as described in any of the first aspects.

[0071] Fifthly, embodiments of this application provide a computer program product that, when run on a terminal device, causes the terminal device to execute the method described in any one of the first aspects above.

[0072] The beneficial effects of the embodiments in this application compared with the prior art are:

[0073] In this embodiment, a multi-layered zonal spatial load model is constructed based on thermoelectric load data under the constraints of distributed energy conditions and thermoelectric balance. The spatial load prediction value is determined according to the multi-layered zonal spatial load model. Since the hierarchical zonal load prediction method considering distributed power sources and thermoelectric coupling is taken into account, the accuracy of the load model used for integrated energy system stability simulation can be improved, and the accuracy of load distribution prediction results can be improved.

[0074] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0075] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this specification. Attached Figure Description

[0076] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0077] Figure 1 This is a schematic flowchart of a user-side thermal and electrical load prediction method provided in an embodiment of this application;

[0078] Figure 2 This is a comparison chart of partial cell load forecasting and adjustment results provided in one embodiment of this application;

[0079] Figure 3 This is a comparison chart of cell load and classified load error before and after adjustment provided in an embodiment of this application;

[0080] Figure 4 This is a comparison diagram of the error between the cell load and the central area load before and after adjustment, provided in an embodiment of this application;

[0081] Figure 5 A schematic diagram of the user-side thermoelectric load prediction device provided in one embodiment of this application;

[0082] Figure 6 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. Detailed Implementation

[0083] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0084] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0085] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0086] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0087] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0088] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0089] In practical planning work, planners need to predict not only the total load but also the specific growth and distribution of the load. This is the main research content of spatial load forecasting. Economic, climatic, environmental, and many other uncertainties affect the load. In particular, due to the popularization of new energy sources and the construction of integrated energy systems, a large number of distributed power sources and thermoelectric coupling devices have been added to the power system. The uncertainties of distributed power sources and thermoelectric balance constitute a complex, multivariate, and nonlinear mapping relationship with the load curve. Therefore, load forecasting not only needs to consider electrical load but also, considering various factors and accurately predicting the integrated energy load, has become the biggest challenge in load forecasting research. With the large-scale adoption of information technology in the power system and various sectors of the national economy, collecting historical load data and related factors has become easier than before. Coupled with continuously improving intelligent computing methods, this provides advantageous tools and research methods for integrated energy load forecasting.

[0090] However, the data on distributed energy sources and user-side thermoelectric loads and their related influencing factors are large in volume, have poor regularity, and are difficult to classify. Furthermore, distributed power sources are highly random and vary significantly across different regions. The coupling characteristics of thermoelectric loads also differ considerably from region to region. How to analyze the characteristics of distributed power sources and thermoelectric coupled loads based on spatial analysis, and how to classify historical data of distributed power sources and thermoelectric loads at different levels and in different regions, is a key challenge in this research.

[0091] Traditional spatial load forecasting models typically employ a two-layer partitioning structure, meaning they only have data on total load and local load. Relying solely on the mutual correction between local and total loads makes it difficult to obtain accurate load distribution forecasts, and land partitioning alone cannot account for the impact of distributed power generation and thermoelectric loads.

[0092] Based on the above problems, this application provides a user-side thermal and electrical load prediction method. By integrating different load division methods and under the constraints of distributed power sources and thermal and electrical balance, a multi-layer partition prediction model is obtained. Through comprehensive adjustment of the prediction results, the average load error is minimized through iteration, resulting in high spatial user electricity consumption prediction accuracy.

[0093] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings and specific implementation methods.

[0094] Figure 1 This is a schematic flowchart of a user-side thermal and electrical load prediction method provided in an embodiment of this application, with reference to... Figure 1 The user-side heat and power load prediction method is described in detail below:

[0095] In step 101, thermoelectric load data is acquired.

[0096] In some embodiments, obtaining thermoelectric load data may include: obtaining raw thermoelectric load data; identifying and correcting abnormal data points in the raw thermoelectric load data using longitudinal comparison and lateral comparison methods to obtain the thermoelectric load data.

[0097] For example, the horizontal comparison method is used to process abnormal data, including: based on the original load data, a smoothed sequence is first obtained by averaging. The difference between the original sequence and the smoothed sequence can be used to obtain an error sequence. Abnormal data points are identified by comparing the values ​​of the error sequence with a set threshold. Then, the corresponding points in the smoothed sequence are directly used to replace the abnormal values ​​to obtain the thermal load data.

[0098] The process can be expressed by the following formula:

[0099]

[0100] Where h1 and h2 represent the indices of two horizontally adjacent points in the load data; L represents the horizontally corrected load data; L origin This represents the load data before horizontal correction; α and β are threshold values, which are generally set to α = β = 0.9 based on practical experience. If the threshold is too small, the data correction sensitivity will be too low, and abnormal data will be missed; if the threshold is too high, it will lead to overcorrection, eliminating some load change characteristic information. Therefore, it should be adjusted according to the situation in practical applications.

[0101] For example, the longitudinal comparison method is used to process outlier data, including: Let L(p,Q) represent the load series sample data, where p = 1, 2, ..., 288 represents 288 moments in a day obtained at five-minute sampling intervals, and Q = 1, 2, ..., N represents the number of sampling days. The mean of the load series sample is calculated according to the following formula.

[0102]

[0103] The variance V(p) at each time point is calculated using the following formula:

[0104]

[0105] If the following condition is met, then the data point is considered a bad data point:

[0106]

[0107] The normal data points in the dataset remain unchanged, while the bad data points are treated with the mean. After making corrections, the thermoelectric load data is obtained.

[0108] In step 102, based on the heat and power load data, a multi-level zonal spatial load model is constructed under the constraints of distributed energy conditions and heat and power balance. The load data of each level in the multi-level zonal spatial load model is divided into total load data, classified load data, central zone load data and small zone load data.

[0109] Specifically, a multi-level zonal spatial load model is constructed under the constraints of distributed energy conditions and thermo-electric balance constraints, including: setting load intervals for thermo-electric load data and performing consistency processing on the load intervals; the load intervals are divided into total load intervals, classified load intervals, central load intervals, and small load intervals; determining the distributed power source condition constraints and thermo-electric coupling load condition constraints; and constructing a multi-level zonal spatial load model based on the load intervals, distributed power source condition constraints, and thermo-electric coupling load condition constraints.

[0110] The setting of the number of intervals for spatial load forecasting includes: the load data of each layer in the spatial load forecasting interval model is the number of intervals, and the most probable value is taken between the upper and lower bounds of the interval. Let t∈{1,2,…,T} be the load classification number, and the data of each layer includes: ① Total load [L - s ,L + s ] and its possible values ​​L s ; ② Total load of type t [L - st ,L + st ] and its possible values ​​L st ③ Central zone load [L] - ma ,L + ma ] and its possible values ​​L ma ④ Community load [L] - sa ,L + sa ] and its possible values ​​L sa .

[0111] Then, load interval consistency processing is performed, including: ensuring that the classified load is consistent with the total load; ensuring that the mid-zone load is consistent with the total load; ensuring that the cell load is consistent with the mid-zone load; and ensuring that the cell load is consistent with the classified load. The following formula is used to ensure that the classified load is consistent with the total load:

[0112]

[0113] Equation (6) above represents the classified load L st The total value should be within the total load range [L] - s ,L + s[Inside]; where Δ(.,.) represents the calculation of the total load after weighting and superimposing the classified loads according to the load curves; L st C represents the total load of type t; st =[C st (0) C st (1) ,…,C st (i) ,…,C st (288) ] represents the 288-point load curve sampled every five minutes on a typical day for type t load, where C st (i) ∈[0,100], C st (i) It is the ratio after dividing by the largest element in the vector as the denominator.

[0114] The following formula ensures that the load in the central zone matches the total load:

[0115]

[0116] Equation (7) above indicates that the total load value of all central zones should be within the total load range; where, L ma (m) represents the load in the middle zone, numbered m; N MA The total load in the central area; α ma-s This refers to the simultaneous rate of load from the central zone to the total load.

[0117] The following formula ensures that the load in the small cell area matches the load in the central cell area:

[0118]

[0119] Equation (8) above indicates that the sum of all load values ​​in any given zone should be within the total load range. - ma ,L + ma Indicates the upper and lower limits of the load in the middle zone; L sa This represents the load of any cell within the central area.

[0120] According to the following formula (9), the cell load is made consistent with the classified load:

[0121]

[0122] Equation (9) above represents the summation value of the load of the t-th cell within the total load range of that class. Where N SA α represents the total load of the cell; sa-s,t is the concurrency rate when summarizing the load of the t-th type of cell.

[0123] Specifically, the constraints on distributed energy resources include: reactive power constraints of wind power generation and operational constraints of energy storage systems;

[0124] The reactive power constraint for wind power generation is:

[0125]

[0126] Among them, P w X represents the active power output of the wind turbine, tanδ is the active power coefficient of the wind turbine, and X... σ X is the sum of the generator stator reactance and rotor reactance. m R is the magnetizing reactance, R is the rotor resistance, and s is the slip.

[0127] The operating constraints of the energy storage system are:

[0128]

[0129] in, and Let t be the maximum and minimum heat storage capacity of the thermal storage system. For different times, the heating temperature and recuperation temperature at the nodes of the energy storage system will change, and the SOC range of the energy storage system will also change to some extent. and These are the maximum and minimum energy storage values ​​of the electrical energy storage system.

[0130] Specifically, the thermoelectric coupling load condition constraints are as follows:

[0131]

[0132] Where, Φ t It is the sum of the system heat load and network heat loss of the thermal system in time period t, obtained through power flow calculation; N h For all centralized heat sources and CHP unit nodes of the system; N e N represents the set of all generator nodes in the system. b P represents the set of all distributed heat pump and electric boiler nodes in the system. t It is the sum of the system power load and network loss in the power system during time period t, which is obtained through power flow calculation; These are the thermal energy and electrical energy emitted by the energy source at node j, respectively, and x j This refers to the system's operating status. and These represent the heat absorption and charging amount of the thermal and electrical energy storage devices at time t, respectively. and η represents the heat release and discharge of the thermal and electrical energy storage devices during time period t, respectively. qdis and η pdis These are the energy release efficiencies of thermal energy storage devices and electrical energy storage devices, respectively.

[0133] For example, the constraints of thermoelectric coupling load conditions mainly consider that all electric and heat load demands are met in real time through cogeneration CHP units, various generators, heat sources, energy storage devices, and the power grid, and that there is no shortage of energy or waste of electrical and heat energy.

[0134] For example, the construction of a multi-level zonal spatial load model includes: after the cell, central zone, and total loads meet the above consistency conditions, the grid load is then corrected based on the adjusted cell load; the grid load is the load after considering the constraints of distributed power source conditions and thermoelectric coupling load conditions; given an arbitrary cell numbered j, its load value is L. sa (j), the sum of all grid loads within the area is denoted as ΣL G Because the grid area is small and the time it takes for the load to reach saturation is very short (a few years), the load growth has a strong step-like nature. Therefore, grid load cannot be corrected as smoothly as in central or small zones. Thus, the adjustment objective is to minimize the error between the grid load and the small zone load, i.e.:

[0135] min|∑L G -L sa (j)| (12)

[0136] On the other hand, distributed power generation constraints and thermo-electric coupling constraints need to be added. Let x(j) be the weighting factor of distributed power generation output and heat load for each cell, and let the heat load be the aforementioned zoned load. Therefore, the hierarchical zoned load considering distributed energy and thermo-electric coupling can be obtained as follows:

[0137] P w +P t =∑(L G (j)-L sa (j))x(j)=∑L(j)x(j) (13)

[0138] Let P(t) be the output of the thermoelectric system; L(t) be the row vector of the zoned load; and X(t) be the thermoelectric load weight corresponding to the zoned load. Then, the hierarchical zoned load model considering distributed energy and thermoelectric coupling can be written in matrix form:

[0139] P(t)=L(t)X(t) (14)

[0140] The above model is a typical overdetermined linear equation system, which can be solved by a comprehensive iterative recursive method to obtain the heat load weights corresponding to the partitions.

[0141] In step 103, the multi-level partitioned spatial load model is initialized.

[0142] Specifically, the initialization of the multi-level partitioned spatial load model includes: verifying the multi-level partitioned spatial load model; initializing the iteration error of the verified multi-level partitioned spatial load model; the iteration error initialization includes: initialization of the iteration forgetting weight matrix, initialization of the covariance matrix and initialization of the partitioned load weights; the forgetting weight matrix, covariance matrix and partitioned load weights are the parameters of the iteration error.

[0143] For example, verifying whether a feasible solution exists includes: verifying whether the cell load can be adjusted to match the classified load; and verifying whether the cell load can be adjusted to match the central area load.

[0144] The process of verifying whether the cell load can be adjusted to match the classified load includes: firstly, calculating the classified load for each type of load using the lower and upper bounds of the cell load interval; the calculation result for the t-th type of load is... and The resulting interval represents the maximum adjustable range of the aggregated load value for this type of cell. Next, this interval is verified against the categorized load interval [L]. - s ,L + s Check if the intersection of [] is empty. If it is not empty, proceed to the next step; otherwise, there is no feasible solution, and the cell load or classified load needs to be corrected and the step is repeated.

[0145] Verifying whether the cell load can be adjusted to match the central zone load includes: for each central zone, calculating the central zone load based on the lower and upper bounds of the cell loads within the central zone. The calculation result for the central zone numbered m is Δ({ΣL MA(m),t -},[C st ]) and Δ({ΣL MA(m),t +},{C st Next, check if the intersection of the resulting interval and the central load interval is empty. If it is not empty, proceed to the next step; otherwise, if there is no feasible solution, the cell load or central load needs to be adjusted, and the step is repeated.

[0146] For example, iteration error initialization includes: calculating E before iteration. ma,sa and E s,sa Initial value. Define the average relative error between the central and small zones:

[0147]

[0148] Where, Δ({k) m,t ΣL MA(m),t +},{C st}) represents the total load value of the cells within this central area; E is the sum of the loads in all central zones. ma,saThe larger the value, the greater the difference between the load in the small area and the load in the central area.

[0149] Define the system-cell load average relative error:

[0150]

[0151] Among them, L st For total load of category t; This is the summative load value for this type of community; E represents the sum of the categorical loads. s,sa The larger the value, the greater the difference between the community load and the category load.

[0152] Among them, the partition load weight initialization is as follows:

[0153] x(j)=1 (17)

[0154] For example, the iterative forgetting weight matrix is ​​initialized using the following formula:

[0155] W(1)=ρ=0.7 (18)

[0156] The forgetting weight matrix at measurement time t is denoted by W(t), where ρ is the forgetting factor. W(t) aims to improve the utilization rate of new data. Since we are concerned with the current topological state, we need to increase the weight of the latest measurement data and decrease the weight of past measurement data. As can be seen from the form of the forgetting weight matrix, when 0 < ρ < 1, the weight of the latest measurement data is 1, while the weight of past measurements decays exponentially, thereby increasing the utilization rate of new measurement data. The larger ρ is chosen, the greater the influence of past measurement results on the current estimator, the slower the estimator changes, and the worse the real-time performance. The smaller ρ is chosen, the faster the forgetting speed of past measurements, the higher the real-time performance, but the more sensitive the estimator is to parameter changes, that is, small parameter changes will lead to abrupt changes in the estimator.

[0157] For example, the covariance matrix is ​​initialized using the following formula:

[0158] Let the covariance matrix K(t) and its inverse K(t) be given. -1 for:

[0159]

[0160] The covariance matrix at time t-1 is obtained as follows:

[0161]

[0162] Write P(t) and L(t) in recursive form as well:

[0163]

[0164] Therefore, the solution to the covariance matrix can be calculated:

[0165] K(t) -1 =ρL(t-1) T W(t-1)L(t-1)+L(t) T L(t) (22)

[0166] Next, construct the recurrence relation for the covariance matrix:

[0167] K(t) -1 -L(t) T L(t)=ρK(t-1) -1 (twenty three)

[0168] In step 104, the multi-level partitioned spatial load model is iteratively solved based on the initial data of the multi-level partitioned spatial load model.

[0169] Specifically, based on the initial data of the multi-level partitioned spatial load model, the multi-level partitioned spatial load model is iteratively solved, including: obtaining the residual of the multi-level partitioned spatial load model based on the initial data of the iteration error of the multi-level partitioned spatial load model; obtaining the objective function of the multi-level partitioned spatial load model based on the residual of the multi-level partitioned spatial load model; minimizing the objective function; minimizing the objective function includes: updating the iteration error and updating the cell load; updating the residual of the multi-level partitioned spatial load model based on the updated iteration error and cell load.

[0170] The process includes updating the iteration error and updating the cell load, which involves: updating the forgetting weight matrix, updating the covariance matrix, and updating the zoning load weights; adjusting the cell load according to the central zone load based on the heat load zoning weights; adjusting the cell load according to the category load based on the heat load zoning weights; and correcting cell loads that exceed the boundaries. After the above correction process of adjusting cell loads according to the central zone load and adjusting cell loads according to the category load, if the cell load exceeds the interval boundary, the cell load is set as the boundary value and will not be adjusted again; the residual is calculated based on the updated data.

[0171] For example, consider the iterative adjustment of hierarchical and zoned load forecasting based on distributed energy and thermoelectric coupling. The purpose of iteration is to continuously adjust the zone load to make E... ma,sa and E s,sa The load is gradually reduced to ensure consistency of the thermoelectric load in the interval model.

[0172] For example, obtain the residual of the zoned integrated heat and power load model: Let the residual of the zoned integrated heat and power load be ε(t):

[0173] ε(t)=E ma,sa (t)+E s,sa(t) (24)

[0174]

[0175]

[0176] For example, the principle of comprehensive iteration is to minimize the sum of squares of the residuals. Therefore, the objective function J(t) can be minimized as follows:

[0177] J(t)=ε(t) T W(t)ε(t) (27)

[0178] For example, minimizing the objective function of a multi-level partitioned spatial load model: To achieve this minimum objective function, taking the first partial derivative of J(t) and setting it to 0, we get:

[0179] 0 = [-L(t)X(t) + P(t)] T W(t) T L(t) (28)

[0180] Further processing yields:

[0181] L(t) T W(t) T K(t)=[L(t) T W(t) T L(t)]X(t) (29)

[0182] Let [L(t)] T W(t) T If L(t) is invertible, then the solution to X(t) is:

[0183] X(t)=[L(t) T W(t) T L(t)] -1 [L(t) T W(t) T P(t)] (30)

[0184] Combining the above K(t) with the solution of the above heat load zoning weights, we can obtain:

[0185] X(t)=K(t)[ρL(t-1) T W(t-1)P(t-1)+L(t) T P(t)] (31)

[0186] For example, the forgetting weight matrix is ​​updated using the following formula:

[0187]

[0188] For example, the covariance matrix is ​​updated using the following formula:

[0189] P(t)=[ρI m (t-1) T W(t-1)I m (t-1)+i m (t) T i m (t)] -1 (33)

[0190] For example, the formula for updating the partition load weight is as follows:

[0191] X(t) = X(t-1) + K(t)L(t) T (P(t)-L(t)X(t-1)) (34)

[0192] For example, the cell load is adjusted according to the heat load zoning weights and the load of each zoning zone.

[0193] For example, the cell load is adjusted according to the heat load zoning weights and the central zone load. For each central zone, the adjustment is performed as follows: the set of cells within the central zone numbered m is denoted as S. MA(m) The number of them is N. SA Calculate the total load of the cells within the central area, and then determine the direction of load adjustment for each cell, which can be divided into the following three cases:

[0194] (1) If the total load value of the cell is within the load range of the central zone, the condition that the cell load is consistent with the central zone load is met, and no adjustment is required.

[0195] (2) If the total load of a cell exceeds the upper limit of the load in its central zone, the cell load should be reduced. Cells whose load has reached the lower limit of their zone cannot have their load reduced further; this zone is designated as a non-adjustable zone, denoted as S. N,MA(m) The number is N SAN The sum of the loads after heat load weighting adjustment is Other cell intervals are called adjustable cells, denoted as S. Y,MA(m) =S MA(m) —S N,MA(m) In the formula, "—" indicates the complement operation. The load adjustment coefficient k of the adjustable sub-area within the central area is defined. MA(m) as follows:

[0196]

[0197] In the formula, L MA(m) The most likely value for the load in the central zone; Δ({ΣL SA},{C st}) represents the sum of the load values ​​of the residential areas within the central zone.

[0198] Get k MA(m) Then, adjust the adjustable cell load according to the following formula: L SAY =L SAY k MA(m) The corrected total cell load value will be equal to L. MA(m) .

[0199] (3) If the total load of a cell is lower than the lower limit of the load in the central zone, the cell load should be increased. Cells that cannot be adjusted are those whose load has reached the upper limit. Other cells that can be adjusted should be adjusted in the same way as in case 2.

[0200] For example, cell loads are adjusted according to load category based on heat load zoning weights. After correcting cell loads for each zone in step 4.6, cell loads need to be adjusted according to load category. For each load category, the adjustment is performed as follows: the set of all cells in the t-th load category is denoted as S. A,t The number is N sa,at Calculating the total load value for this type of community and then determining the direction of load adjustment also involves three scenarios:

[0201] (1) The total load value of the cell is within the total load range of this category, and meets the consistency condition between the cell load and the category load, so no adjustment is required.

[0202] (2) If the total load of a cell exceeds the upper limit of the total load for that category, the cell load should be reduced. Cells of category t that have reached the lower limit of the interval and cannot be adjusted are denoted as S. N,t The number is N SAN,st The sum of the loads after heat load weighting adjustment is Other adjustable cells are S Y,t =S A,t -S N,t Define the adjustment factor k for the load of the t-th adjustable cell. t as follows:

[0203]

[0204] In the formula: L st The most likely value for the classification load; α sa-s Let t be the simultaneous rate of load from the t-th cell to the corresponding class of load; Let k be the sum of the loads of the t-th type of cell. t Then, adjust the adjustable cell as follows: L SAY,st =L SAY, st k t .

[0205] (3) If the total load value of a cell is lower than the lower limit of the classified load interval, the cell load should be increased. Unadjustable cells are those that have reached the upper limit of the load interval; adjustable cells should be adjusted using the same method as in case 2.

[0206] For example, correct the load of a cell that exceeds the boundary. After the above steps, if the cell load exceeds the interval boundary, the cell load is set as the boundary value and will not be adjusted again. Finally, calculate the residual ε(t) of the combined thermal and electrical load of the zone.

[0207] In step 105, if the iterative solution result meets the preset conditions, the spatial load prediction value is output.

[0208] In this process, the calculated residual ε(t) is compared with the result ε(t-1) of the previous iteration to determine whether the preset conditions are met. If they are met, the iteration ends; otherwise, the process jumps to step 104 to continue the iteration.

[0209] Specifically, the preset conditions are:

[0210] max(|ε(t)-ε(t-1)|)<E max (37)

[0211] Among them, E max The threshold is set to 0.001 in the example, and t represents the number of iterations.

[0212] It is evident that the user-side thermoelectric load prediction method proposed in this invention can integrate different division methods and simultaneously consider distributed power sources and thermoelectric coupling in a multi-layered zoned load prediction model, including levels such as total power, total load, classified power, classified load, distributed power sources, and thermoelectric equipment. Through comprehensive adjustment of the prediction results, the average load error is minimized through iteration, achieving high-precision load prediction for various zoned loads. Furthermore, the comprehensive iterative adjustment method can further improve the accuracy of load prediction.

[0213] All the above steps can be implemented using C programming language and Visual Studio. For greater convenience and efficiency, MATLAB is generally used. An example and analysis will be presented below.

[0214] For example, we will analyze a spatial load forecasting case of a medium-sized city with a long-term planned area of ​​80 km². 2 The load and land use classifications are as follows: Number 1 is industrial land; numbers 12 to 17 are residential, commercial / warehouse, education / health, transportation / postal / telecommunications, agriculture / animal husbandry / fishery, and other land uses, respectively. The multi-level zoning of the example is shown in Table 1.

[0215] Table 1 Multi-level partitioning details

[0216]

[0217] The load forecast results for each floor in a certain year of the example were comprehensively adjusted, and the iteration error is shown in Table 2. During the iteration process, the residual ε (%) gradually decreased, with the largest decrease in the first iteration, followed by a gradual decrease thereafter. After 6 iterations, the error tended to be constant, and the iteration ended.

[0218] Table 2 Errors during the iterative adjustment process

[0219]

[0220] The adjustment algorithm corrected the load of 1772 cells for the year. The changes in some cells before and after the adjustment are shown in Table 3 and... Figure 2 As shown in the table, the load adjustments in each community are all carried out within their respective ranges. For example, the load of community "C19-1240" was adjusted from 240kW to 225kW, which is within the range [204, 264].

[0221] Table 3 Load Adjustment for Some Residential Areas (kW)

[0222]

[0223]

[0224] The errors in cell load and classified load before and after adjustment are shown in Table 4 and Figure 3 The table shows that before adjustment, the average relative error E between the cell load and the classified load was... s,sa The error rate reached 7.19%, with the absolute values ​​of errors for categories 3 and 5 exceeding 10%, and the total load values ​​for all categories of cells exceeding the range of total load for that category. (Adjusted E) s,sa The load rate dropped to 0.52%, the maximum absolute value of the error dropped to 0.98%, and the total load values ​​of all types of communities returned to the range of total load for that type, achieving consistency.

[0225] Table 4 Errors in Cell Load and Classified Load Before and After Adjustment

[0226]

[0227] Before and after the adjustment, the load difference between the small area and the central area is shown in the figure. Figure 4 After calculation and comprehensive adjustments, E ma,sa The error rate decreased from 9.48% to 1.6%, and the maximum error decreased from 17.19% to 3.73%. After adjustment, the total load values ​​of all sub-districts within the central zone returned to the corresponding central zone intervals, and the prediction error met the requirements of practical applications. This demonstrates the accuracy and effectiveness of the interval power consumption prediction model and iterative adjustment method proposed in this paper.

[0228] It should be understood that the sequence number of each step above does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0229] Corresponding to the user-side thermal and electrical load prediction method in the above embodiments, Figure 5 A structural block diagram of the user-side thermoelectric load prediction method apparatus provided in the embodiments of this application is shown. For ease of explanation, only the parts related to the embodiments of this application are shown.

[0230] See Figure 5 The user-side thermoelectric load prediction method apparatus in this application embodiment may include an acquisition module 201, a construction module 202, an initialization module 203, a solution module 204, and an output module 205.

[0231] The system includes: an acquisition module 201 for acquiring thermal and electrical load data; a construction module 202 for constructing a multi-level zonal spatial load model based on the thermal and electrical load data under distributed energy constraints and thermal and electrical balance constraints; the load data of each level in the multi-level zonal spatial load model is divided into total load data, classified load data, central zone load data, and small zone load data; an initialization module 203 for initializing the multi-level zonal spatial load model; a solution module 204 for iteratively solving the multi-level zonal spatial load model based on the initialization data; and an output module 205 for outputting the spatial load prediction value if the iterative solution result meets preset conditions.

[0232] It should be noted that the information interaction and execution process between the above-mentioned devices are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0233] This application also provides a terminal device, see [link to relevant documentation] Figure 6 The terminal device 300 may include at least one processor 310 and a memory 320, wherein the memory 320 stores a computer program 321 that can run on the at least one processor 310, and the processor 310 executes the computer program to implement the steps in any of the above-described method embodiments, for example... Figure 1 Steps 101 to 105 in the illustrated embodiment. Alternatively, when the processor 310 executes the computer program, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 5 The functions of modules 201 to 205 are shown.

[0234] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 320 and executed by processor 310 to complete this application. The one or more modules / units may be a series of computer program segments capable of performing specific functions, which describe the execution process of the computer program in terminal device 300.

[0235] Those skilled in the art will understand that Figure 6 This is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than shown, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0236] The processor 310 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0237] The memory 320 can be an internal storage unit of the terminal device or an external storage device, such as a plug-in hard drive, a smart media card (SMC), a secure digital card (SD), or a flash card. The memory 320 is used to store the computer program and other programs and data required by the terminal device. The memory 320 can also be used to temporarily store data that has been output or will be output.

[0238] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0239] The user-side thermoelectric load prediction method provided in this application embodiment can be applied to terminal devices such as computers, tablets, laptops, netbooks, and personal digital assistants (PDAs). This application embodiment does not impose any restrictions on the specific type of terminal device.

[0240] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the various embodiments of the user-side thermoelectric load prediction method described above.

[0241] This application provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps in the various embodiments of the user-side thermoelectric load prediction method described above.

[0242] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0243] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0244] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0245] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0246] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0247] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for predicting user-side thermal and electrical loads, characterized in that, include: Acquire thermal load data; Based on the aforementioned thermal and electrical load data, a multi-layered zonal spatial load model is constructed under distributed energy constraints and thermal and electrical balance constraints. The load data of each layer in the multi-layered zonal spatial load model is divided into total load data, classified load data, central zone load data, and small zone load data. Initialize the multi-layered partitioned spatial load model; Based on the initial data of the multi-level partitioned spatial load model, the multi-level partitioned spatial load model is solved iteratively. If the iterative solution result meets the preset conditions, the spatial load prediction value is output; The construction of a multi-layered zonal spatial load model under distributed energy constraints and heat and power balance constraints includes: The heat and power load data is configured with load intervals, which are divided into total load intervals, category load intervals, mid-zone load intervals, and sub-zone load intervals. The load range is subjected to consistency processing; Determine the constraints of distributed power source conditions and thermoelectric coupling load conditions; A multi-layered spatial load model is constructed based on the load range, the distributed power source constraints, and the thermoelectric coupling load constraints. The thermoelectric coupling load condition constraint is as follows: Where, Φ t It is a thermal system t The sum of the system heat load and network heat loss for a given period is obtained through power flow calculations; N h For all centralized heat sources and CHP unit nodes of the system; N e N represents the set of all generator nodes in the system. b This refers to the set of all distributed heat pump and electric boiler nodes in the system. P t It is the power system in t The sum of the system power load and network loss for a given time period is obtained through power flow calculation. , p t j These are nodes j The energy source generates heat and electricity. x j Its running status; and p t chr Thermal and electrical energy storage devices are respectively located in t The amount of heat absorbed and the amount of charge generated in a short period of time; and p t dis Thermal and electrical energy storage devices are respectively located in t The amount of heat and discharge during the period η qdis and η pdis These are the energy release efficiencies of thermal energy storage devices and electrical energy storage devices, respectively.

2. The user-side heat and power load prediction method as described in claim 1, characterized in that, The acquisition of thermoelectric load data includes: Obtain raw data on thermal and electrical loads; Abnormal data points in the original thermoelectric load data are identified and corrected by longitudinal comparison and lateral comparison methods to obtain the thermoelectric load data.

3. The user-side heat and power load prediction method as described in claim 1, characterized in that, The step of setting load ranges for the processed raw thermoelectric load data includes: set up t {1,2,…, T } represents the load classification number. Data at each level includes: total load range [ L - s , L + s ], No. t Total load range [ L - st , L + st ], Central load zone [ L - ma , L + ma ] and the load range of the community[ L - sa , L + sa ]; The consistency processing of the load range includes: pass Make the classification load L st The total value should be within the total load range [ L - s , L + s [Inside]; where Δ(.,.) represents the calculation of the total load after the classified loads are weighted and superimposed according to the load curves; L st Indicates the first t The classification load described above; C st =[ C st (0) , C st (1) , …, C st (i) ,…, C st (288) ] indicates the first t The 288-point load curves, sampled every five minutes on a typical day of the classified load, are described in the class. C st (i) [0,100], C st (i) It is the ratio after dividing by the largest element in the vector as the denominator; pass The total load value of all the aforementioned central zones should be within the total load range; wherein, L ma ( m ) indicates that the number is m The aforementioned central load; α ma-s The simultaneity rate from the central load to the total load; pass The summation value of the load of any of the aforementioned cells within the central zone is within the load interval of the central zone; L - ma , L + ma Indicates the upper and lower limits of the load in the central zone; L sa For any cell within the central area; pass Make the first t The total load value of the cells described in this class is within the total load range of this class; α sa-s,t For the first t Simultaneity rate when aggregating the load of different types of cells; The consistency processing can obtain the optimal value for each partition load range.

4. The user-side heat and power load prediction method as described in claim 1, characterized in that, The distributed energy constraints include: reactive power constraints for wind power generation and operational constraints for energy storage systems. The reactive power constraint for wind power generation is: in, The wind turbine is active and generates power. The active power coefficient of the wind turbine X σ This is the sum of the generator stator reactance and the rotor reactance. X m For the magnetizing reactance, R For rotor resistance, s For worsening; The operating constraints of the energy storage system are: in, and for t The maximum and minimum heat storage capacity of the thermal energy storage system varies at different times. The supply and recovery temperatures at the nodes of the energy storage system will change accordingly. SOC The scope will also change to some extent; and These are the maximum and minimum energy storage values ​​of the electrical energy storage system.

5. The user-side thermal and power load prediction method as described in claim 1, characterized in that, The initialization of the multi-layered partitioned spatial load model includes: Verify the multi-layered spatial load model; The iterative error initialization is performed on the verified multi-level partitioned spatial load model; the iterative error initialization includes: iterative forgetting weight matrix initialization, covariance matrix initialization, and partitioned load weight initialization; the forgetting weight matrix, the covariance matrix, and the partitioned load weight are the parameters of the iterative error.

6. The user-side thermal and power load prediction method as described in claim 5, characterized in that, The iterative solution of the multi-level partitioned spatial load model based on the initialization data of the multi-level partitioned spatial load model includes: Based on the iterative error initialization data of the multi-level partitioned spatial load model, the residual of the multi-level partitioned spatial load model is obtained; Based on the residuals of the multi-level partitioned spatial load model, the objective function of the multi-level partitioned spatial load model is obtained. Minimize the objective function; Minimizing the objective function includes: Update the iteration error and update the cell load; Based on the updated iteration error and cell load, the residuals of the multi-level partitioned spatial load model are updated.

7. The user-side thermal and electrical load prediction method as described in claim 1, characterized in that, The preset conditions are: in, E max The threshold value is set to 0.001 in the example, which is a pre-defined threshold. t This represents the number of iterations.

8. A terminal device, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the user-side thermoelectric load prediction method as described in any one of claims 1 to 7.

Images