Multi-element electric energy substitution interval prediction method and system based on time-delay MGM model
Through the multivariate electrical energy substitution interval prediction method based on the time-delay MGM model, the problem of inaccurate prediction of the potential of electrical energy substitution is solved, and the accurate prediction and upper and lower boundary provision of the influence of multiple factors of the amount of electrical energy substitution are achieved, which improves the accuracy of the prediction and Validity.
Patent Information
- Application Number
- CN202411049547.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-11-19
AI Technical Summary
The existing prediction model of electric energy substitution potential cannot accurately describe the trend of electric energy substitution under the influence of multiple factors, resulting in the problem of inaccurate prediction.
The multivariate electrical energy substitution interval prediction method based on the time-delay MGM model is adopted. By obtaining the data and historical data of the influencing factor of the electric energy substitution, the interval grayscale sequence and the time-delay MGM model are calculated, and the upper and lower bounds of the electric energy substitution amount are predicted, and the particle swarm algorithm is used to adjust it. Time delay parameters to improve prediction accuracy.
It effectively reflects the complex variable relationship between multiple factors of electric energy substitution, improves the accuracy and effectiveness of forecasting trends of electric energy substitution in multiple fields, provides the upper and lower boundaries of electric energy substitution, and helps make scientific decisions.
Smart Images

Figure CN118982111A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of green energy technology, and more specifically to a multivariate electric energy substitution interval prediction method and system based on a time-lag MGM model. Background Art
[0002] For a long time, the coal-based energy consumption system and high-carbon industrial structure have caused my country's total carbon emissions and intensity to show a "double high" phenomenon. For this reason, my country has proposed a "dual carbon" goal, planning to achieve carbon neutrality in 2060, reduce residents' dependence on traditional energy in production and life, and improve the current situation of serious atmospheric pollution. Electricity substitution is a consumption method that uses electricity to replace fossil energy in the terminal energy consumption link, which is one of the most effective measures to achieve the "dual carbon" goal. my country's electricity substitution work, with the core content of "replacing coal with electricity, replacing oil with electricity, and electricity coming from afar, and it is clean electricity", has received widespread attention from the society.
[0003] With the help of algorithms such as least squares support vector machines, neural networks and hybrid models, researchers are committed to establishing accurate and reliable prediction models to evaluate the feasibility of electric energy substitution. These studies are based on various statistical data and use different algorithms and evaluation methods to determine the electric energy substitution potential in different regions, providing a scientific basis for local energy planning. Due to problems such as sampling errors, reporting errors or time differences, the actual variables used in the prediction of electric energy substitution potential cannot be accurately described by real numbers. At present, most prediction models are point prediction models, which cannot solve the problem of inaccurate predictions caused by data errors.
[0004] Therefore, it is an urgent problem for those skilled in the art to propose a multivariate electric energy substitution interval prediction method and system based on a time-delay MGM model to solve the difficulties existing in the prior art. Summary of the invention
[0005] In view of this, the present invention provides a multivariate electric energy substitution interval prediction method and system based on the time-lag MGM model, which can effectively reflect the complex variable relationship of multiple factors of electric energy substitution and solve the problem of trend prediction of electric energy substitution in multiple fields under the influence of multiple factors.
[0006] In order to achieve the above object, the present invention provides the following technical solutions:
[0007] A multivariate electric energy substitution interval prediction method based on a time-lag MGM model comprises the following steps:
[0008] Obtain data on factors affecting electric energy substitution and historical data on electric energy substitution;
[0009] Based on the data of factors influencing electric energy substitution and the historical data of electric energy substitution, the interval grayscale sequence is calculated, and the interval grayscale sequence includes a kernel sequence and a grayscale sequence;
[0010] Different time-delay MGM models are calculated according to the kernel sequence, grayscale sequence and set time-delay parameters;
[0011] The prediction values of the kernel sequence and the gray sequence are obtained respectively through different time-delay MGM models, and the prediction errors of the kernel sequence and the gray sequence are calculated according to the prediction values;
[0012] The prediction error is judged, and based on the judgment result, it is decided whether to restore the kernel sequence and grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution.
[0013] In the above method, optionally, the data on factors affecting electric energy substitution include regional economic factors, technical factors and environmental factors.
[0014] In the above method, optionally, different time-lag MGM models include a time-lag MGM model established based on time-lag parameters and a kernel sequence and a time-lag MGM model established based on time-lag parameters and a grayscale sequence.
[0015] In the above method, optionally, the nuclear sequence calculation formula is:
[0016] definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,...,n. The calculation formula for the kth element in the sequence is:
[0017]
[0018] In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy replacement at the k-th moment;
[0019]
[0020] j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula of the pth element in the sequence is:
[0021]
[0022] In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
[0023] In the above method, optionally, the grayscale sequence calculation formula in S2 is:
[0024] definition is the grayscale sequence of the electric energy replacement in the ith field at time k=1,2,…,n. The calculation formula of the kth element in the sequence is:
[0025]
[0026] In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is,
[0027] definition is the grayscale sequence of the j-th multivariate electric energy substitution influencing factor at time k=1,2,…,n. The calculation formula of the k-th element in the sequence is:
[0028]
[0029] In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is,
[0030] In the above method, optionally, the time-delay MGM model established based on the time-delay parameter and the kernel sequence is:
[0031] in,
[0032]
[0033] are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. is the element corresponding to the first moment in the first-order accumulation generation sequence of the electric energy substitution quantity core sequence, It is the first-order accumulation sequence generated by the t-period time-delay kernel sequence of the factors related to electric energy substitution at time k.
[0034] In the above method, optionally, the time-lag MGM model established based on the time-lag parameter and the grayscale sequence is:
[0035]
[0036] in are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. It is the element corresponding to the first moment in the sequence generated by the first-order accumulation of the grayscale sequence of the electric energy substitution amount. It is the sequence generated by first-order accumulation of the t-period time-lagged grayscale sequence of the factors related to electric energy substitution at time k.
[0037] In the above method, optionally, the specific steps of judging the prediction error are:
[0038] Determine whether the prediction error of the kernel sequence and the grayscale sequence is greater than the preset threshold. If so, adjust the lag parameter based on the particle swarm algorithm, output the adjusted lag parameter and return to the third step of calculating the prediction error, recalculate the prediction error until the prediction error of the kernel sequence and the grayscale sequence is less than the preset threshold, restore the kernel sequence and the grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution; otherwise, directly restore the kernel sequence and the grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution.
[0039] A multivariate electric energy substitution interval prediction system based on a time-lag MGM model, using any of the above-mentioned multivariate electric energy substitution interval prediction methods based on a time-lag MGM model, comprising a data acquisition module, an interval grayscale sequence calculation module, a model calculation module, a prediction value acquisition module, and a prediction interval acquisition module connected in sequence; wherein,
[0040] A data acquisition module is used to acquire data on factors affecting electric energy substitution and historical data on electric energy substitution amount;
[0041] An interval grayscale sequence calculation module is used to calculate an interval grayscale sequence based on the data of factors affecting electric energy substitution and the historical data of electric energy substitution, wherein the interval grayscale sequence includes a kernel sequence and a grayscale sequence;
[0042] A model calculation module is used to calculate different time-delay MGM models according to the kernel sequence, the gray sequence and the set time-delay parameters;
[0043] A prediction value acquisition module is used to obtain the prediction values of the nuclear sequence and the gray sequence respectively through different time-delay MGM models, and calculate the prediction errors of the nuclear sequence and the gray sequence according to the prediction values;
[0044] The prediction interval acquisition module is used to judge the prediction error and decide whether to restore the kernel sequence and grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution according to the judgment result.
[0045] In the above system, optionally, the data on factors affecting electric energy substitution acquired in the data acquisition module include regional economic factors, technical factors and environmental factors.
[0046] In the above system, optionally, the kernel sequence calculation formula in the interval grayscale sequence calculation module is:
[0047] definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,…,n. The calculation formula for the kth element in the sequence is:
[0048]
[0049] In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy replacement at the k-th moment;
[0050]
[0051] j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula of the pth element in the sequence is:
[0052]
[0053] In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
[0054] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements any of the above-mentioned multivariate electric energy substitution interval prediction methods based on a time-delay MGM model.
[0055] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a multivariate electric energy substitution interval prediction method and system based on a time-delay MGM model, which has the following beneficial effects:
[0056] (1) The present invention is based on the MGM (1, m, N) grey prediction model, which not only considers the impact of economic, environmental, technical and political factors on the amount of multi-dimensional electric energy substitution, but also considers the mutual influence of electric energy substitution in different fields. It can effectively reflect the complex variable relationship of multiple factors of electric energy substitution and solve the trend prediction problem of electric energy substitution in multiple fields under the influence of multiple factors;
[0057] (2) The present invention introduces a time lag parameter into the MGM (1, m, N) model and estimates the time lag parameter based on a particle swarm algorithm, fully considering the time lag effect of factors affecting electric energy substitution, thereby improving the effectiveness and rationality of model prediction;
[0058] (3) In order to solve the problem that a large amount of actual data cannot be accurately described by real numbers due to sampling errors, collection conditions and other factors, the present invention establishes a gray prediction model based on interval gray numbers to predict the upper and lower limits of the amount of multi-electric energy substitution, better handle uncertainty and provide more comprehensive information; the early prediction and evaluation of the potential for multi-electric energy substitution can help governments, enterprises and users make decisions, and therefore has practical significance. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0060] Figure 1 A flow chart of a multivariate electric energy substitution interval prediction method based on a time-lag MGM model provided by the present invention;
[0061] Figure 2 A structural diagram of a multivariate electric energy substitution interval prediction system based on a time-delay MGM model provided by the present invention. DETAILED DESCRIPTION
[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0063] Interval prediction can provide upper and lower bounds for the change of electric energy substitution, better handle uncertainty and provide more comprehensive information. The grey system theory was proposed by Professor Deng Julong in the early 1980s, aiming to provide a new solution for "small sample" and "poor information" uncertainty systems with "partial information known and partial information unknown". Many studies have studied and improved the theory and applied it to various fields. Among them, the time-delay MGM model effectively solves the simulation and prediction problems of multiple system characteristic behavior variables under the influence of multiple related factors, and has great application prospects.
[0064] Reference Figure 1 As shown, the present invention discloses a multivariate electric energy substitution interval prediction method based on a time-lag MGM model, comprising the following steps:
[0065] S1. Obtaining data on factors affecting electric energy substitution and historical data on electric energy substitution;
[0066] S2. Calculate the interval grayscale sequence based on the data of factors affecting electric energy substitution and the historical data of electric energy substitution, where the interval grayscale sequence includes a kernel sequence and a grayscale sequence;
[0067] S3, calculating different time-delay MGM models according to the kernel sequence, grayscale sequence and set time-delay parameters;
[0068] S4, obtaining the predicted values of the kernel sequence and the grayscale sequence respectively through the different time-delay MGM models in S3, and calculating the prediction errors of the kernel sequence and the grayscale sequence according to the predicted values;
[0069] S5. Judge the prediction error in S4, and decide whether to restore the kernel sequence and the grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution according to the judgment result.
[0070] Specifically, the time-lag MGM model is the MGM (1, m, N) grey prediction model.
[0071] Further, S1, obtaining data on factors affecting electric energy substitution and historical data on electric energy substitution:
[0072] 1.1 Quantifying factors affecting regional economy. The level of regional economic development in year t can be measured by regional GDP x1(t), per capita GDP x6(t), per capita electricity consumption x3(t), coal-electricity price ratio x4(t) and oil-electricity price ratio x5(t).
[0073] 1.2 Quantify the regional electricity consumption. Technological progress is mainly manifested in scientific and technological innovations such as replacing oil with electricity, replacing steam with electricity, and replacing coal with electricity. This paper uses the electrification index to reflect the regional energy technology level, and its expression is:
[0074]
[0075] Where x6(t) is the quantified regional energy technology level in year t; E p (t) is the electricity consumption in year t; E f (t) is the equivalent electrical energy consumption of all energy sources in the region in year t.
[0076] 1.3 Quantification of environmental protection factors. Environmental protection needs are specifically manifested in the continuous optimization of the energy structure under the dual carbon background, the gradual replacement of fossil energy consumption by electricity consumption, and the reduction of atmospheric pollutants generated during fossil energy consumption. As the proportion of fossil energy gradually decreases during the process of electricity replacement, the emission of polluting gases will also change significantly. This patent uses carbon emissions and polluting gas emissions as indicators to reflect environmental protection factors, and the expression is:
[0077] x7(t)=C e (t)
[0078] Where x7(t) is the quantified regional environmental protection level in year t; C e (t) is the regional carbon emissions in year t.
[0079] Further, S2, construct interval grayscale sequence:
[0080] The interval grayscale sequence includes the kernel sequence and the grayscale sequence. This part constructs the interval gray number sequence for the historical data of the electric energy substitution amount and the influencing factors respectively.
[0081] 2.1 Definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,…,n. The calculation formula for the kth element in the sequence is:
[0082]
[0083] In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; It is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
[0084] 2.2 j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula of the pth element in the sequence is:
[0085]
[0086] In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
[0087] 2.3 Definition is the grayscale sequence of the electric energy replacement in the ith field at time k=1,2,…,n. The calculation formula of the kth element in the sequence is:
[0088]
[0089] In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is,
[0090] 2.4 Definition is the grayscale sequence of the j-th multivariate electric energy substitution influencing factor at time k=1,2,...,n, and the calculation formula of the k-th element in the sequence is:
[0091]
[0092] In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is,
[0093] S3 includes S301: Establishing a time-delay MGM (1, m, N) model based on kernel sequence
[0094] 1) The time lag parameter t of the jth multivariate electricity substitution related factor j If the jth multivariate electric energy substitution related factor sequence is known, then t j The time-delay kernel sequence is:
[0095]
[0096] 2) Calculation The first-order cumulative generation sequence of in
[0097]
[0098]
[0099] 3) Calculate the i-th electric energy substitution core sequence The first-order cumulative generation sequence of in
[0100]
[0101]
[0102] 4) Calculation The nearest mean of the generated sequence in
[0103]
[0104]
[0105] 5) Calculate the parameter sequence β i =(a i1 ,...,a iM ,b i2 ,...,b iN ) T ,i=1,2,…,M, the calculation method is: β i =(P T P -1 ) -1 P T Q i ;
[0106] in
[0107]
[0108]
[0109] 6) Solve the time response of the MGM(1,m,N) model based on the kernel sequence
[0110]
[0111] in, are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. is the element corresponding to the first moment in the first-order accumulation generation sequence of the electric energy substitution quantity core sequence, It is the first-order accumulation sequence generated by the t-period time-delay kernel sequence of the factors related to electric energy substitution at time k.
[0112] Step S302: Establishing a time-delay MGM (1, m, N) model based on grayscale sequence
[0113] 1) The time lag parameter t of the jth multivariate electricity substitution related factor j If the jth multivariate electric energy substitution related factor is known, then t j The grayscale sequence with time lag is:
[0114]
[0115] 2) Calculation The first-order cumulative generation sequence of in
[0116]
[0117]
[0118] 3) Calculate the grayscale sequence of the i-th electric energy substitution The first-order cumulative generation sequence of in
[0119]
[0120]
[0121] 4) Calculation The nearest mean of the generated sequence in
[0122]
[0123]
[0124] 5) Calculate the parameter sequence γ i =(c i1 ,...,c im ,d i2 ,...,diN ) T ,i=1,2,…,M, the calculation method is: γ i =(R T R -1 ) -1 R T S i ;
[0125] in:
[0126]
[0127]
[0128] 6) Solve the time response of the MGM(1,m,N) model based on grayscale sequence
[0129]
[0130] in are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. It is the element corresponding to the first moment in the sequence generated by the first-order accumulation of the grayscale sequence of the electric energy substitution amount. It is the sequence generated by first-order accumulation of the t-period time-lagged grayscale sequence of the factors related to electric energy substitution at time k.
[0131] Furthermore, the predicted value of the nuclear sequence in S4 is:
[0132]
[0133] in, is the prediction result of the multivariate electric energy substitution kernel sequence at time k+1; is the prediction result of the core of the electric energy substitution amount in the ith field at time k+1; The prediction result of the first-order accumulation generation sequence of the multivariate electric energy substitution quantity kernel sequence at time k+1; The prediction result of the first-order cumulative generation sequence value corresponding to the electric energy substitution quantity core of the i-th field at the k+1 time;
[0134] The prediction error of the kernel sequence is:
[0135]
[0136] Among them, M represents the number of types of fields where the electric energy substitution amount to be predicted is located, n represents the number of moments of the electric energy substitution amount, represents the prediction result of the core of the electric energy substitution amount in the i-th field at time k in the model, is the actual value of the core of the electric energy substitution in the ith field at time k.
[0137] Furthermore, the predicted value of the grayscale sequence in S4 is:
[0138]
[0139] in, is the predicted value sequence of the grayscale of the multivariate electric energy substitution at time k+1;
[0140] The prediction error of the grayscale sequence is:
[0141]
[0142] in, represents the prediction result of the grayscale of the electric energy substitution amount in the i-th field at time k in the model, is the actual value of the grayscale of the electric energy substitution in the i-th field at time k.
[0143] Furthermore, the specific steps of judging the prediction error in S4 in S5 are:
[0144] Determine whether the prediction error of the kernel sequence and the grayscale sequence is greater than the preset threshold. If so, adjust the lag parameter based on the particle swarm algorithm, output the adjusted lag parameter and return to S3-S5 until the prediction error of the kernel sequence and the grayscale sequence is less than the preset threshold, and restore the upper and lower bounds of the prediction values of the kernel sequence and the grayscale sequence; otherwise, directly restore the upper and lower bounds of the prediction values of the kernel sequence and the grayscale sequence.
[0145] Optimizing the time-delay parameters of factors related to electric energy substitution based on particle swarm optimization:
[0146] Using the particle swarm optimization algorithm, the time lag parameter t of the relevant factors of electric energy substitution is obtained. j The optimal value of .
[0147] 5.1 Randomly generate a group of N-dimensional particles in the search space, and the particle positions correspond to the values of the grayscale sequence time-delay parameters of the electric energy substitution related factors. Initialize the position and speed of the particle group.
[0148] 5.2 Substitute the initialized time lag parameters into S3-S5, calculate the prediction errors of the multivariate electric energy substitution kernel sequence and the grayscale sequence respectively, and take the average of these two errors as the fitness of each particle, which is expressed as follows:
[0149]
[0150] 5.3 Update the individual optimal solution and global optimal solution of the particle according to the fitness value. Compare the fitness value of the current particle with the fitness value of the individual optimal solution and the global optimal solution, and update the individual optimal solution and the global optimal solution of the particle.
[0151] 5.4 Based on the information of the individual optimal solution and the global optimal solution, calculate the new velocity and position of the particle. Specifically, update according to the following formula:
[0152] v i (t+1)=w·v i (t)+c1·rand()·[p best,i (t)-x i (t)]+c2·rand()·[g best (t)-x i (t)]
[0153] x i (t+1)=x i (t)+int(v i (t+1))
[0154] Among them, v i (t) and x i (t) represent the velocity and position of the ith particle at the tth iteration, respectively, and p best , i(t) represents the individual optimal solution of the i-th particle at the t-th iteration, g best (t) represents the global optimal solution at the tth iteration, w, c1, c2 and rand() represent the inertia weight, self-cognition factor, social cognition factor and random number function respectively; int(v i (t+1)) represents the i (t+1) is rounded down to the nearest integer. Through the above formula, the speed and position of the particle are adjusted according to the information of the individual optimal solution and the global optimal solution.
[0155] 5.5 Repeat steps 5.2-5.5 until the termination condition is met. The termination condition can be reaching the preset maximum number of iterations or the global optimal solution has no change after multiple consecutive iterations. After the termination condition is met, the algorithm ends and outputs the global optimal solution.
[0156] Restore the simulated prediction values of the kernel sequence and grayscale sequence:
[0157] The upper and lower bounds of the predicted values of the restored kernel sequence and grayscale sequence are as follows:
[0158] By restoring the simulated predicted values of the kernel sequence and the grayscale sequence, the upper and lower bounds of the interval gray number are calculated:
[0159]
[0160]
[0161] Among them, l up With l down are the upper and lower boundaries of the interval gray number background domain, Represents the prediction result of the core of the electric energy substitution amount in the i-th field at time k in the model.
[0162] Reference Figure 2 As shown, a multivariate electric energy substitution interval prediction system based on a time-lag MGM model, a multivariate electric energy substitution interval prediction method based on a time-lag MGM model according to any one of the above items, comprising a data acquisition module, an interval grayscale sequence calculation module, a model calculation module, a prediction value acquisition module, and a prediction interval acquisition module connected in sequence; wherein,
[0163] A data acquisition module is used to acquire data on factors affecting electric energy substitution and historical data on electric energy substitution amount;
[0164] An interval grayscale sequence calculation module is used to calculate an interval grayscale sequence based on the data of factors affecting electric energy substitution and the historical data of electric energy substitution, wherein the interval grayscale sequence includes a kernel sequence and a grayscale sequence;
[0165] A model calculation module is used to calculate different time-delay MGM models according to the kernel sequence, the gray sequence and the set time-delay parameters;
[0166] A prediction value acquisition module is used to obtain the prediction values of the nuclear sequence and the gray sequence respectively through different time-delay MGM models, and calculate the prediction errors of the nuclear sequence and the gray sequence according to the prediction values;
[0167] The prediction interval acquisition module is used to judge the prediction error and decide whether to restore the kernel sequence and grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution according to the judgment result.
[0168] Furthermore, the data on factors affecting electric energy substitution acquired in the data acquisition module include regional economic factors, technical factors and environmental factors.
[0169] Furthermore, the kernel sequence calculation formula in the interval grayscale sequence calculation module is:
[0170] definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,...,n. The calculation formula for the kth element in the sequence is:
[0171]
[0172] In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy replacement at the k-th moment;
[0173]
[0174] j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula for the pth element in the sequence is:
[0175]
[0176] In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
[0177] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements any of the above-mentioned multivariate electric energy substitution interval prediction methods based on a time-delay MGM model.
[0178] In a specific embodiment, in order to verify the feasibility of the model proposed in this patent, a multivariate electric energy substitution interval prediction model is established using data from 11 years from 2007 to 2017, and the electric boiler capacity and heat pump capacity in the region from 2018 to 2022 are predicted. The prediction results of the electric boiler and heat pump capacity are shown in Table 1 and Table 2. It can be seen from the table that the relative error of the upper and lower bounds of the electric boiler capacity of the model proposed in this patent is less than 10%. The results show that the prediction accuracy of the model is high and the effect is good.
[0179] Table 1
[0180]
[0181] Table 2
[0182]
[0183] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0184] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A multivariate electric energy substitution interval prediction method based on a time-delay MGM model, characterized in that: The following steps are involved: Obtain data on factors affecting electric energy substitution and historical data on electric energy substitution; Based on the data of factors influencing electric energy substitution and the historical data of electric energy substitution, the interval grayscale sequence is calculated, and the interval grayscale sequence includes a kernel sequence and a grayscale sequence; Different time-delay MGM models are calculated according to the kernel sequence, grayscale sequence and set time-delay parameters; The prediction values of the kernel sequence and the gray sequence are obtained respectively through different time-delay MGM models, and the prediction errors of the kernel sequence and the gray sequence are calculated according to the prediction values; The prediction error is judged, and based on the judgment result, it is decided whether to restore the kernel sequence and grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution.
2. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 1 is characterized in that: The data on factors affecting electric energy substitution include regional economic factors, technical factors and environmental factors.
3. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 1 is characterized in that: The nuclear sequence calculation formula is: definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,...,n. The calculation formula for the kth element in the sequence is: In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy replacement at the k-th moment; j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula of the pth element in the sequence is: In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
4. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 1 is characterized in that: The grayscale sequence calculation formula is: definition is the grayscale sequence of the electric energy replacement in the ith field at time k=1,2,…,n. The calculation formula of the kth element in the sequence is: In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is, definition is the grayscale sequence of the j-th multivariate electric energy substitution influencing factor at time k=1,2,…,n. The calculation formula of the k-th element in the sequence is: In the formula, l up With l down are the upper and lower boundaries of the interval gray number background domain, that is, 5. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 1 is characterized in that: Different time-lag MGM models include the time-lag MGM model based on time-lag parameters and kernel sequence and the time-lag MGM model based on time-lag parameters and grayscale sequence.
6. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 5 is characterized in that: The time-delay MGM model established based on the time-delay parameter and kernel sequence is: in, are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. is the element corresponding to the first moment in the first-order accumulation generation sequence of the electric energy substitution quantity core sequence, It is the first-order accumulation sequence generated by the t-period time-delay kernel sequence of the factors related to electric energy substitution at time k.
7. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 5 is characterized in that: The time-lag MGM model established based on the time-lag parameter and grayscale sequence is: in are differential equation parameter matrices, The first-order accumulation of the electric energy substitution at time k+1 generates a sequence prediction value. It is the element corresponding to the first moment in the sequence generated by the first-order accumulation of the grayscale sequence of the electric energy substitution amount. It is the sequence generated by first-order accumulation of the t-period time-lagged grayscale sequence of the factors related to electric energy substitution at time k.
8. The multivariate electric energy substitution interval prediction method based on the time-delay MGM model according to claim 1 is characterized in that: The specific steps for judging the prediction error are: Determine whether the prediction error of the kernel sequence and the gray sequence is greater than the preset threshold. If so, adjust the lag parameter based on the particle swarm algorithm, output the adjusted lag parameter and return to the third step of calculating the prediction error, recalculate the prediction error until the prediction error of the kernel sequence and the gray sequence is less than the preset threshold, and restore the kernel sequence and the gray sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution; Otherwise, the kernel sequence and grayscale sequence are directly restored to obtain the upper and lower bounds of the predicted value of the electric quantity substitute.
9. A multivariate electric energy substitution interval prediction system based on a time-delay MGM model, characterized in that: A multivariate electric energy substitution interval prediction method based on a time-delay MGM model as described in any one of claims 1 to 8 is applied, comprising a data acquisition module, an interval grayscale sequence calculation module, a model calculation module, a prediction value acquisition module, and a prediction interval acquisition module connected in sequence; wherein, A data acquisition module is used to acquire data on factors affecting electric energy substitution and historical data on electric energy substitution amount; An interval grayscale sequence calculation module is used to calculate an interval grayscale sequence based on the data of factors affecting electric energy substitution and the historical data of electric energy substitution, wherein the interval grayscale sequence includes a kernel sequence and a grayscale sequence; A model calculation module is used to calculate different time-delay MGM models according to the kernel sequence, the gray sequence and the set time-delay parameters; A prediction value acquisition module is used to obtain the prediction values of the nuclear sequence and the gray sequence respectively through different time-delay MGM models, and calculate the prediction errors of the nuclear sequence and the gray sequence according to the prediction values; The prediction interval acquisition module is used to judge the prediction error and decide whether to restore the kernel sequence and grayscale sequence to obtain the upper and lower bounds of the predicted value of the electric quantity substitution according to the judgment result.
10. The multivariate electric energy substitution interval prediction system based on the time-delay MGM model according to claim 9 is characterized in that: The data on factors affecting electric energy substitution obtained in the data acquisition module include regional economic factors, technical factors and environmental protection factors.
11. The multivariate electric energy substitution interval prediction system based on the time-delay MGM model according to claim 9, characterized in that: The calculation formula of the kernel sequence in the interval grayscale sequence calculation module is: definition is the core sequence of the electric energy substitution in the ith field at time k=1,2,…,n. The calculation formula for the kth element in the sequence is: In the formula, is the upper bound of the value of the i-th multivariate electric energy substitution at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy replacement at the k-th moment; j=1,2,…,N is defined as the core sequence of the jth multivariate electric energy substitution influencing factor at time p=-l,…,-1,0,1,…,s. The calculation formula of the pth element in the sequence is: In the formula, is the upper bound of the value of the j-th multivariate electricity substitution influencing factor at the k-th moment; is the lower bound of the value of the i-th multivariate electric energy substitution at the k-th moment.
12. A computer-readable storage medium, characterized in that: A computer program is stored on the storage medium, and when the computer program is executed by the processor, a multivariate electric energy substitution interval prediction method based on a time-delay MGM model as described in any one of claims 1 to 8 is implemented.
