Intelligent electric energy distribution device for industrial park

Abstract

The invention discloses an intelligent electric energy distribution device for an industrial park. The intelligent electric energy distribution device comprises a data acquisition unit, a cloud power plant unit and an intelligent regulation and control unit. The data acquisition unit records the total generating capacity of modes of connecting industrial park power transmission line power supply end green energy and coal-fired power generation and the like and the real-time power consumption of the industrial park, and the data acquisition unit and the real-time power consumption upload measured data to the cloud power plant unit; the cloud power plant unit constructs an LSTM prediction model, learns the data, and predicts a difference value between the total power generation amount and the power consumption amount. A TTAO optimization algorithm is constructed, and the optimal electricity price and the optimal distribution of electric power resources are found; and the intelligent regulation and control unit adjusts the electricity price and the electric power resource according to the optimal electricity price found by the cloud power plant unit and the optimal electric power resource distribution. Compared with the prior art, electricity price adjustment and power resource allocation are carried out according to the total power generation amount of the power supply end of the power transmission line connected with the industrial park, and the problems that a virtual power plant mechanism is not flexible in regulation and control and resources cannot be allocated can be solved.

Description

Technical Field

[0001] The present invention belongs to automatic energy distribution technology, relates to an intelligent electric energy distribution technology, and specifically relates to an industrial park electric energy intelligent distribution device. Background Art

[0002] In recent years, my country's industry has developed rapidly, with a large scale and a sound system, and many enterprises have gathered to form industrial parks.

[0003] However, due to the large number of factories in industrial parks and high electricity consumption, although my country has vigorously developed new energy and new energy represented by wind, solar and storage have been connected to the power grid on a large scale, the volatility of power supply has increased significantly, the problem of power grid balance has intensified, and the insufficient regulation capacity on the traditional power supply side often leads to power outages in industrial parks, which in turn affects the production of major factories in the park.

[0004] At present, in order to prevent the power consumption from being less than the total power generation, most industrial parks optimize their equipment and use equipment with lower energy consumption. Although this method can reduce power consumption, the equipment is expensive and the production cost is rising, which is not conducive to the economic development of enterprises. Some enterprises also build energy storage systems to store excess electricity when the power demand is low and release electricity when the demand is high to balance the supply and demand of electricity. This method requires a larger floor area and has high requirements for the factory area and environment, which is not suitable for the development of some small and medium-sized enterprises. In addition, the government has invested heavily in enterprises, encouraged enterprises to transform, optimize the energy structure, and provide appropriate electricity subsidies. However, this method is highly dependent on the outside, which is not conducive to the development of enterprises themselves, and there will be a waste of power resources.

[0005] Therefore, an intelligent energy distribution technology is needed that can monitor the electricity consumption of the industrial park and adjust the electricity price and allocate power resources according to the total power generation at the power supply end of the transmission line connected to the industrial park. Summary of the invention

[0006] Purpose of the invention: In response to the problems pointed out in the background technology, the present invention provides an industrial park electric energy intelligent distribution device to monitor the power consumption of the industrial park, and adjust the electricity price and allocate power resources according to the total power generation at the power supply end of the transmission line connected to the industrial park.

[0007] Technical solution: The present invention discloses an industrial park electric energy intelligent distribution device, comprising an interconnected data acquisition unit, a cloud power plant unit and an intelligent control unit;

[0008] The data collection unit measures the total power generation of the energy at the power supply end of the power transmission line connected to the industrial park, as well as the real-time power consumption of the industrial park, and uploads the collection results to the cloud power plant unit;

[0009] The cloud power plant unit receives the information collected by the data acquisition unit, learns the collected data, and predicts the difference ΔP between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the next moment by building an LSTM prediction model. When the difference ΔP exceeds the threshold ΔP surplus or less than the threshold ΔP shortage When ΔP is used, a triangle topology aggregation optimization algorithm TTAO is constructed to find the best price baseline and send the result to the intelligent control unit; ΔP shortage It refers to the threshold value when the power generation is less than the power consumption and a gap is about to occur; ΔP surplus Refers to the threshold at which power generation is considered to be greater than power consumption, which will result in power waste;

[0010] The intelligent control unit is responsible for processing the instructions issued by the cloud power plant unit, implementing tiered electricity prices, and allocating power resources.

[0011] Furthermore, the cloud power plant unit predicts the difference between the total power generation at the power supply end of the transmission line connected to the industrial park and the power consumption of the industrial park at the next moment. The specific steps are as follows:

[0012] 2.1: For the data obtained by the data collection unit, use the 3sigma principle and maximum likelihood estimation method to deal with outliers and missing values;

[0013] 2.2: Calculate the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park:

[0014] ΔP=(P c +P w +k s ·ε s )-P i

[0015] Among them, ΔP is the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park; P c is the coal-fired power generation; P w Wind power generation; k s is a constant, indicating the maximum possible power generation of solar power generation; ε s The probability of sunny days in the industrial park area; P i For electricity consumption in industrial parks.

[0016] 2.3: According to the structural relationship of the LSTM prediction model, the cell state value at the tth moment is obtained, that is, the predicted difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the tth moment is:

[0017]

[0018] In the formula, S t , S t-1 are the cell state values ​​at time t and time t-1 respectively;

[0019] Finally, the expression of the hidden layer state value at the tth moment is:

[0020]

[0021] In the formula, O t is the output value in the feature extraction process:

[0022] O t =σ(W o ·h t-1 +U o ·x t +b o )

[0023] Where W o The weight coefficient in the feature extraction process, b o Bias term in the feature extraction process.

[0024] Further, the difference ΔP is compared. When the difference is within the range [ΔP surplus , +∞], it is believed that the total power generation at the power supply end of the transmission line connecting the industrial park is much greater than the power consumption of the industrial park, and there is a waste of power resources. Lowering the electricity price baseline can stimulate the power consumption of enterprises and reasonably allocate the excess power resources to achieve the maximum utilization of resources; when the difference range is [-ΔP shortage ,ΔP surplus ], it is considered that the total power generation at the power supply end of the transmission line connected to the industrial park just meets the power consumption of the industrial park and no adjustment is required; when the difference range is [-∞,ΔP shortage ], it is believed that the total power generation at the power supply end of the transmission line connecting the industrial park is less than the power consumption of the industrial park, and there is a gap in power resources. It is necessary to raise the electricity price baseline to allow enterprises to save electricity and increase the power generation equipment at the power supply end of the transmission line to meet the electricity demand.

[0025] Furthermore, according to the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park, the improved triangle topology aggregation optimization algorithm TTAO is selected to find the best electricity price baseline and realize the regulation of power resources:

[0026] Step 3.1: Construct the electricity price objective function:

[0027] M=B+∑ i (k i ·(MB)·I i )i

[0028] Where M is the actual electricity price; i represents different electricity price intervals, B is the electricity price baseline; k i I is the electricity price adjustment coefficient; i is the interval index;

[0029] Step 3.2: The triangle topology aggregation optimization algorithm TTAO is based on similar triangles. With iterations, new vertices are continuously generated in the search space and used to form similar triangles of different sizes. In the algorithm, each triangle is regarded as a basic evolutionary unit with four bodies, namely, three vertices of the triangle and one internal random vertex;

[0030] Step 3.3: Population initialization. Each vertex in the triangular topology unit represents a search agent. In the triangular topology aggregation optimization algorithm TTAO, the number of solutions N of the electricity price baseline is divided into N / 3 triangular topology units; the additional electricity price baseline solutions are randomly generated in the search space. In the initialization phase, N / 3 agents are randomly generated in the feasible region. The mathematical expression generated by each individual is:

[0031]

[0032] in, represents the first electricity price baseline in the i-th triangle topology unit, where i is a positive integer between 1 and N / 3, and r0 represents a random number between [0, 1]; and are the lower and upper bounds of the variable;

[0033] Step 3.4: In the triangular topological unit formation stage, a new direction vector with a length of l*f is located using the solution of the first valence baseline as the starting vertex in the spherical coordinate system, and is converted to the ordinary coordinate system through trigonometric functions to form the second vertex, which is the solution of the second valence baseline; then, the direction vector with a length of l*f is generated and rotated counterclockwise by π / 3, and then the solution of the third valence baseline is obtained through coordinate system transformation. The three solutions constitute a triangular topological unit, and the expressions of these solutions are written as:

[0034]

[0035] Where l represents the size of the triangular topological unit, and represents the direction vectors of the other two edges guided by the first point, θ=[θ1,…,θ D ] and θ j (j=1,…,D) is a random number between [0,π];

[0036] Each group of triangular topological units is internally aggregated into a fourth vertex, which is the fourth solution to the electricity price baseline that meets the conditions, formed in a linearly weighted manner to use individual information, defined as:

[0037]

[0038] Where r1, r2, r3 are random numbers between [0, 1] and r1 + r2 + r3 = 1, and the fourth search agent is located in each triangular topological unit;

[0039] Step 3.5: In the global aggregation phase, information about the solutions of the excellent electricity price baselines in different triangular units composed of different electricity price baselines is collected, and a new feasible solution is created; information interaction occurs between the best electricity price baseline solution in each triangular topological unit and the best solution in any randomly selected unit set, and there is a linear combination of different weights between each dimensional variable of the two positive electricity price baseline solutions. The new electricity price baseline solution is generated by connecting the two better electricity price baseline solutions, which can be expressed as:

[0040]

[0041] Where r4 is a random number between [0,1], represents the new baseline solution for electricity prices in the ith unit, and represents the optimal electricity price baseline solution for unit i and a set of electricity price baseline solutions randomly selected at the tth iteration;

[0042] The greedy strategy is used to update the optimal agent. The mathematical expression is:

[0043]

[0044] In the formula, represents the suboptimal electricity price baseline solution for the i-th unit, represents the optimal position in the ith unit at the t+1th iteration; Indicates bringing in The magnitude of the direction vector at this solution is Indicates bringing in The magnitude of the direction vector at this solution; Indicates bringing in The magnitude of the direction vector at this solution;

[0045] Step 3.6: local polymerization;

[0046] Step 3.6.1: A triangular topology is temporarily formed between the updated optimal or suboptimal electricity price baseline solution and the two electricity price baseline solutions in the group with good fitness values, that is, a group of electricity price baseline solutions. Each group is re-searched in a certain local area to realize the utilization of each topological triangle unit. The calculation formula of the new electricity price baseline solution is:

[0047]

[0048] Among them, α is expressed as: represents the second new electricity price baseline solution at the t+1th iteration in the i-th unit;

[0049] Step 3.6.2: After aggregation, compare the fitness values ​​of the two electricity price baseline solutions before and after local mining to determine the update of the position; if the new solution is better than the original solution, update the position, otherwise do not perform the update. The mathematical expression of the position of the corresponding electricity price baseline solution is as follows:

[0050]

[0051] in, Indicates bringing in The magnitude of the direction vector is Indicates bringing in The magnitude of the direction vector;

[0052] Step 3.7: Use gradient descent method for iterative optimization:

[0053]

[0054] Among them, Y old is the solution of the current electricity price baseline; Y new is the solution of the updated electricity price baseline; η is the learning rate, i.e., the step size, a positive constant that controls the magnitude of the update; is the objective function M in Y old The gradient at , expressed as a vector, points in the direction of the fastest growth of the function;

[0055] Step 3.8: When the number of iterations reaches the maximum or the fitness value meets the requirements, exit the loop and output the optimal electricity price baseline, otherwise return to step 3.5 for a new round of iteration.

[0056] Furthermore, in step 3.4, in order to improve the search space of the algorithm, the formula for updating the size l of the triangular topological unit is as follows:

[0057]

[0058] Among them, t represents the current number of iterations; T represents the maximum number of iterations; l0 is the initial value of l; is a constant that controls the growth rate of l, and l decreases as the number of iterations increases.

[0059] Furthermore, the specific process of the intelligent control unit to achieve the allocation of power resources is:

[0060] When the difference between the total power generation at the power supply end of the industrial park transmission line and the power consumption of the industrial park is within [ΔP surplus ,+∞], the excess power resources are allocated to areas with insufficient power resources, or the electric energy is converted into other energy. When the power consumption peaks, other energy is converted into electric energy and sent to the power grid; when the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park is between [-∞,ΔP shortage ] When power is insufficient, electricity is transferred from other areas with abundant power resources, or additional equipment is added at the power supply end of the transmission line.

[0061] Beneficial effects:

[0062] The present invention adjusts the electricity price benchmark by predicting the difference between the total power generation at the power supply end of the industrial park transmission line and the power consumption of the industrial park at the next moment, thereby adjusting the electricity price, effectively avoiding the problem that the power consumption of the industrial park during peak hours is less than the power generation, and there is excess power during valley hours; secondly, the present invention uses the improved triangular topology aggregation optimization algorithm TTAO to find the best electricity price baseline, the algorithm is faster and more accurate, and can make suggestions for electricity price adjustment to relevant staff at the first time, greatly saving manpower. In addition, the present invention can optimize the power consumption structure of the industrial park to a certain extent through the adjustment of electricity prices, which helps to ensure the output of the factory area and the development of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a schematic diagram of a device unit according to an embodiment of the present invention;

[0064] Figure 2 is a flow chart of a method according to an embodiment of the present invention;

[0065] Figure 3 A data diagram showing the total regional power generation, the power consumption of the industrial park, their difference, and the difference after regulation according to an embodiment of the present invention; Figure 4 A comparison chart of the difference between the average power generation and power consumption of the original industrial park in the four seasons and the difference after the power price adjustment in the embodiment of the present invention; Figure 5 This is a comparison chart of the electricity prices for the four seasons in the industrial park and the electricity prices after regulation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The specific technical solutions of the present invention are further described in detail below.

[0067] The invention discloses an intelligent electric energy distribution device for an industrial park, comprising a data acquisition unit, a cloud power plant unit and an intelligent control unit which are interconnected.

[0068] Data acquisition unit: The data acquisition unit is mainly composed of capacity sensors and load sensors. The capacity sensors are mainly used to measure the total power generation of hydropower, wind power and other energy sources at the power supply end of the transmission line connected to the industrial park; the load sensors are mainly used to measure the real-time power consumption of the industrial park and upload the measured results to the cloud power plant.

[0069] Cloud power plant unit: The cloud power plant unit is mainly responsible for receiving the information collected by the data acquisition unit, learning the collected data, and predicting the difference ΔP between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the next moment by building an LSTM prediction model.

[0070] When the difference ΔP exceeds the threshold ΔP surplus or less than the threshold ΔP shortage When the price is adjusted, a triangle topology aggregation optimization algorithm TTAO is constructed to find the best price baseline and send the result to the intelligent control unit.

[0071] Intelligent control unit: The intelligent control unit is responsible for processing the instructions issued by the cloud power plant unit, implementing tiered electricity prices, and realizing the allocation of power resources.

[0072] The cloud power plant unit predicts the difference between the total power generation at the power supply end of the transmission line connected to the industrial park and the power consumption of the industrial park at the next moment. The specific steps are as follows:

[0073] 2.1 For the data obtained by the data collection unit, the 3 sigma principle and maximum likelihood estimation method are used to deal with outliers and missing values.

[0074] 2.2 Calculate the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park by the formula:

[0075] ΔP=(P c +P w +k s ·ε s )-P i

[0076] Among them, ΔP is the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park; P c is the coal-fired power generation; P w Wind power generation; k s is a constant, indicating the maximum possible power generation of solar power generation; ε sThe probability of sunny days in the industrial park area; P i For electricity consumption in industrial parks.

[0077] 2.3 According to the structural relationship of LSTM, the cell state value at the tth moment is obtained, that is, the predicted difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the tth moment is:

[0078]

[0079] In the formula, S t , S t-1 are the cell state values ​​at time t and time t-1 respectively.

[0080] Finally, the expression of the hidden layer state value at time t is obtained as follows:

[0081]

[0082] In the formula, O t is the output value in the feature extraction process:

[0083] O t =σ(W o ·h t-1 +U o ·x t +b o )

[0084] Where W o The weight coefficient in the feature extraction process, b o Bias term in the feature extraction process.

[0085] Compare the size of the difference. When the difference is within the range [ΔP surplus , +∞], it can be considered that the total power generation at the power supply end of the transmission line connecting the industrial park is much greater than the power consumption of the industrial park, and there is a waste of power resources. It is necessary to lower the electricity price baseline to stimulate the power consumption of enterprises and reasonably allocate the excess power resources to achieve the maximum utilization of resources.

[0086] When the difference range is [-ΔP shortage ,ΔP surplus ], it is considered that the total power generation at the power supply end of the transmission line connected to the industrial park just meets the power consumption of the industrial park and no adjustment is required.

[0087] When the difference range is [-∞,ΔP shortage ], it is believed that the total power generation at the power supply end of the transmission line connecting the industrial park is less than the power consumption of the industrial park, and there is a gap in power resources. At this time, it is necessary to raise the electricity price baseline to allow enterprises to save electricity and increase the power generation equipment at the power supply end of the transmission line to meet the electricity demand.

[0088] According to the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park, the improved TTAO optimization algorithm is used to find the best electricity price baseline and realize the regulation of power resources:

[0089] 3.1 Constructing the electricity price objective function:

[0090] M=B+Σ i (k i ·(MB)·I i )i

[0091] Where M is the actual electricity price; i represents different electricity price intervals, B is the electricity price baseline; k i I is the electricity price adjustment coefficient; i is the interval index; The specific electricity price level and the corresponding difference range need to be determined based on the actual situation of the electricity market, policy regulations and demand fluctuations.

[0092] 3.2TTAO algorithm (Triangle Topology Aggregation Optimization Algorithm) is based on similar triangles. With iterations, new vertices are continuously generated in the search space and used to form similar triangles of different sizes. In the algorithm, each triangle is regarded as a basic evolutionary unit with four bodies, namely the three vertices of the triangle and an internal random vertex.

[0093] 3.3 Population initialization: Each vertex in the triangular topological unit represents a search agent. In the TTAO algorithm, the number of solutions to the electricity price baseline, N, is divided into N / 3 triangular topological units. The solutions to the additional electricity price baseline are randomly generated in the search space. In the initialization phase, N / 3 agents are randomly generated in the feasible region, and the mathematical expression for each individual generation is:

[0094]

[0095] in, represents the first electricity price baseline in the i-th triangle topology unit, and i is a positive integer between 1 and N / 3. r0 represents a random number between [0, 1]; and are the lower and upper bounds of the variable.

[0096] 3.4 Triangular topological unit formation stage

[0097] 3.4.1 A new direction vector of length l*f is located using the solution of the first valence baseline as the starting vertex in the spherical coordinate system and converted to the ordinary coordinate system through trigonometric functions to form the second vertex, which is the solution of the second valence baseline. Next, the direction vector of length l*f is generated and rotated counterclockwise by π / 3, and then the solution of the third valence baseline is obtained through coordinate system transformation. The three solutions constitute a triangular topological unit. The expressions of these solutions can be written as:

[0098]

[0099] Where l represents the size of the triangle topological unit.

[0100] 3.4.2 To improve the search space of the algorithm, the formula for updating l is as follows:

[0101]

[0102] Among them, t represents the current number of iterations; T represents the maximum number of iterations; l0 is the initial value of l, which is a constant that controls the growth rate of l; l decreases as the number of iterations increases.

[0103] 3.4.3 and Represent the direction vectors of the other two edges directed by the first point:

[0104]

[0105] where θ=[θ1,…,θ D ] and θ j (j=1,…,D) is a random number between [0,π].

[0106] 3.4.4 Each group of triangular topological units is internally aggregated into a fourth vertex, which is the fourth solution to the electricity price baseline that meets the conditions. This point is formed in a linearly weighted manner to use individual information and is defined as:

[0107]

[0108] Where r1, r2, r3 are random numbers between [0, 1] and r1 + r2 + r3 = 1. Therefore, the fourth search agent is located in each triangular topological unit.

[0109] 3.5 General (global) aggregation stage

[0110] 3.5.1 In this stage, information about the solutions of excellent price baselines in different triangular units composed of different price baselines is collected and a new feasible solution is created. Information interaction occurs between the best price baseline solution in each triangular topological unit and the best solution in any randomly selected unit set. There is a linear combination with different weights between each dimensional variable of the two positive price baseline solutions. The new price baseline solution is generated by connecting the better two price baseline solutions, which is mathematically expressed as:

[0111]

[0112] Where r4 is a random number between [0,1], represents the new baseline solution for electricity prices in the ith unit, and represents the optimal electricity price baseline solution for unit i and the solution of a set of randomly selected electricity price baselines at the tth iteration.

[0113] 3.5.2 Use the greedy strategy to update the optimal agent. The mathematical expression is:

[0114]

[0115] In the formula, represents the suboptimal electricity price baseline solution at the i-th iteration, represents the optimal position in the ith unit at the t+1th iteration; Indicates bringing in The magnitude of the direction vector at this solution is Indicates bringing in The magnitude of the direction vector at this solution; Indicates bringing in The magnitude of the direction vector at this solution.

[0116] 3.6 Local Aggregation

[0117] 3.6.1 In this stage, the triangular topology units are locally aggregated. After the previous stage, a triangular topology, that is, a group of price baseline solutions, is temporarily formed between the updated optimal or suboptimal price baseline solution and the two price baseline solutions in the group with good fitness values. In this case, the topology is not necessarily an equilateral triangle. Based on the motion vector difference formed by the optimal and suboptimal price baseline solutions, the position of the optimal price baseline solution is disturbed (in terms of direction and step size) in the local area. Therefore, each group is re-searched within a certain local area to achieve the utilization of each topological triangular unit. The calculation formula for the new price baseline solution is:

[0118]

[0119] It represents the second new electricity price baseline solution at the t+1th iteration in the i-th unit. The calculation formula of α is:

[0120]

[0121] 3.6.2 The purpose of using a suboptimal electricity price baseline solution is to prevent the optimal electricity price baseline solution from falling into a local extreme value. After aggregation, the guide point of the temporary triangle unit should be guaranteed to be optimal within a set of electricity price baseline solutions. In order to make the convergence develop in the most practical direction, the fitness values ​​of the two electricity price baseline solutions before and after local mining are compared to determine the update of the position. If the new solution is better than the original solution, the position is updated, otherwise the update is not performed. The mathematical expression of the position of the corresponding electricity price baseline solution is as follows:

[0122]

[0123] in, Indicates bringing in The magnitude of the direction vector is Indicates bringing in The magnitude of the direction vector.

[0124] 3.7 To improve the computational efficiency of the algorithm, the gradient descent method is used here:

[0125]

[0126] Y old is the solution (vector) of the current electricity price baseline; Y new is the solution to the updated electricity price baseline; η is the learning rate (step size), a positive constant that controls the magnitude of the update; is the objective function M in Y old The gradient at , expressed as a vector, points in the direction of fastest growth of the function.

[0127] 3.8 When the number of iterations reaches the maximum or the fitness value meets the requirements, exit the loop, otherwise return to 3.5.2 for a new round of iterations.

[0128] When the difference between the total power generation at the power supply end of the industrial park transmission line and the power consumption of the industrial park is within [ΔP surplus ,+∞], the power resources are too abundant, and the excess power resources can be deployed to the areas with insufficient power resources, or the electric energy can be converted into other energy. For example, when the power consumption is low at night, the excess power can be converted into water potential energy. During the peak power consumption in the daytime, the water potential energy is converted into electric energy and sent to the power grid; when the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park is in [-∞,ΔP shortage ] When power is scarce, electricity can be transferred from other areas with abundant power resources, or additional equipment can be added at the power supply end of the transmission line to increase power generation to meet demand.

[0129] In view of the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park, the improved triangular topology aggregation optimization algorithm TTAO is selected to find the best electricity price baseline, and an industrial park is selected for experiment:

[0130] like Figure 3 , is the total power generation of the region, the power consumption of the industrial park, and their difference and the difference after regulation. It is not difficult to find from the figure that after the model prediction, the difference prediction between the regional power generation and power consumption is more accurate.

[0131] like Figure 4 , is the difference between the average power generation and power consumption of the original industrial park in the region throughout the year and the difference after electricity price regulation; Figure 5 , which are the electricity prices of the industrial park in the region throughout the year and the adjusted electricity prices. Combining the two figures, the electricity price adjusted by the model can reduce the difference between power generation and industrial park power consumption, further avoiding the waste of power resources and improving the economic efficiency of power grid operation.

[0132] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent power distribution device for an industrial park, characterized in that: It includes interconnected data acquisition units, cloud power plant units and intelligent control units; The data collection unit measures the total power generation of the energy at the power supply end of the power transmission line connected to the industrial park, as well as the real-time power consumption of the industrial park, and uploads the collection results to the cloud power plant unit; The cloud power plant unit receives the information collected by the data acquisition unit, learns the collected data, and predicts the difference ΔP between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the next moment by building an LSTM prediction model. When the difference ΔP exceeds the threshold ΔP surplus or less than the threshold ΔP shortage When the triangle topology aggregation optimization algorithm TTAO is constructed, the best price baseline is found and the result is sent to the intelligent control unit; ΔP shortage It refers to the threshold value when the power generation is less than the power consumption and a gap is about to occur; ΔP surplus Refers to the threshold at which power generation is considered to be greater than power consumption, which will result in power waste; The intelligent control unit is responsible for processing the instructions issued by the cloud power plant unit, implementing tiered electricity prices, and allocating power resources.

2. The intelligent power distribution device for industrial parks according to claim 1 is characterized in that: The cloud power plant unit predicts the difference between the total power generation at the power supply end of the transmission line connected to the industrial park and the power consumption of the industrial park at the next moment. The specific steps are as follows: 2.1: For the data obtained by the data collection unit, use the 3sigma principle and maximum likelihood estimation method to deal with outliers and missing values; 2.2: Calculate the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park: ΔP=(P c +P w +k s ·ε s )-P i Among them, ΔP is the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park; P c is the coal-fired power generation; P w Wind power generation; k s is a constant, indicating the maximum possible power generation of solar power generation; ε s The probability of sunny days in the industrial park area; P i For electricity consumption in industrial parks. 2.3: According to the structural relationship of the LSTM prediction model, the cell state value at the tth moment is obtained, that is, the predicted difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park at the tth moment is: In the formula, S t , S t-1 are the cell state values ​​at time t and time t-1 respectively; Finally, the expression of the hidden layer state value at the tth moment is: In the formula, O t is the output value in the feature extraction process: The t =σ(W o ·h t-1 +U o ·x t +b o ) Where W o The weight coefficient in the feature extraction process, b o Bias term in the feature extraction process.

3. The intelligent power distribution device for industrial parks according to claim 2 is characterized in that: Compare the size of the difference ΔP. When the difference is within the range [ΔP surplus , +∞], it is believed that the total power generation at the power supply end of the transmission line connecting the industrial park is much greater than the power consumption of the industrial park, and there is a waste of power resources. Lowering the electricity price baseline can stimulate the power consumption of enterprises and reasonably allocate the excess power resources to achieve the maximum utilization of resources; when the difference range is [-ΔP shortage ,ΔP surplus ], it is considered that the total power generation at the power supply end of the transmission line connected to the industrial park just meets the power consumption of the industrial park and no adjustment is required; when the difference range is [-∞,ΔP shortage ], it is believed that the total power generation at the power supply end of the transmission line connecting the industrial park is less than the power consumption of the industrial park, and there is a gap in power resources. It is necessary to raise the electricity price baseline to allow enterprises to save electricity and increase the power generation equipment at the power supply end of the transmission line to meet the electricity demand.

4. The intelligent power distribution device for industrial parks according to claim 3 is characterized in that: According to the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park, the improved triangle topology aggregation optimization algorithm TTAO is selected to find the best electricity price baseline and realize the regulation of power resources: Step 3.1: Construct the electricity price objective function: M=B+Σ i (k i ·(M-B)·I i )i Where M is the actual electricity price; i represents different electricity price intervals, B is the electricity price baseline; k i I is the electricity price adjustment coefficient; i is the interval index; Step 3.2: The triangle topology aggregation optimization algorithm TTAO is based on similar triangles. With iterations, new vertices are continuously generated in the search space and used to form similar triangles of different sizes. In the algorithm, each triangle is regarded as a basic evolutionary unit with four bodies, namely, three vertices of the triangle and one internal random vertex; Step 3.3: Population initialization. Each vertex in the triangular topology unit represents a search agent. In the triangular topology aggregation optimization algorithm TTAO, the number of solutions N of the electricity price baseline is divided into N / 3 triangular topology units; the additional electricity price baseline solutions are randomly generated in the search space. In the initialization phase, N / 3 agents are randomly generated in the feasible region. The mathematical expression generated by each individual is: in, represents the first electricity price baseline in the i-th triangle topology unit, where i is a positive integer between 1 and N / 3, and r0 represents a random number between [0, 1]; and are the lower and upper bounds of the variable; Step 3.4: In the triangular topological unit formation stage, a new direction vector with a length of l*f is located using the solution of the first valence baseline as the starting vertex in the spherical coordinate system, and is converted to the ordinary coordinate system through trigonometric functions to form the second vertex, which is the solution of the second valence baseline; then, the direction vector with a length of l*f is generated and rotated counterclockwise by π / 3, and then the solution of the third valence baseline is obtained through coordinate system transformation. The three solutions constitute a triangular topological unit, and the expressions of these solutions are written as: Where l represents the size of the triangular topological unit, and represents the direction vectors of the other two edges guided by the first point, θ=[θ1,…,θ D ] and θ j (j=1,…,D) is a random number between [0,π]; Each group of triangular topological units is aggregated internally into a fourth vertex, which is the fourth solution to the electricity price baseline that meets the conditions, formed in a linearly weighted manner to use individual information, defined as: Where r1, r2, r3 are random numbers between [0, 1] and r1 + r2 + r3 = 1, and the fourth search agent is located in each triangular topological unit; Step 3.5: In the global aggregation phase, information about the solutions of the excellent electricity price baselines in different triangular units composed of different electricity price baselines is collected, and a new feasible solution is created; information interaction occurs between the best electricity price baseline solution in each triangular topological unit and the best solution in any randomly selected unit set, and there is a linear combination of different weights between each dimensional variable of the two positive electricity price baseline solutions. The new electricity price baseline solution is generated by connecting the two better electricity price baseline solutions, which can be expressed as: Where r4 is a random number between [0,1], represents the new baseline solution for electricity prices in the ith unit, and represents the optimal electricity price baseline solution for unit i and a set of electricity price baseline solutions randomly selected at the tth iteration; The greedy strategy is used to update the optimal agent. The mathematical expression is: In the formula, represents the suboptimal electricity price baseline solution for the i-th unit, represents the optimal position in the ith unit at the t+1th iteration; Indicates bringing in The magnitude of the direction vector at this solution is Indicates bringing in The magnitude of the direction vector at this solution; Indicates bringing in The magnitude of the direction vector at this solution; Step 3.6: local polymerization; Step 3.6.1: A triangular topology is temporarily formed between the updated optimal or suboptimal electricity price baseline solution and the two electricity price baseline solutions in the group with good fitness values, that is, a group of electricity price baseline solutions. Each group is re-searched in a certain local area to realize the utilization of each topological triangle unit. The calculation formula of the new electricity price baseline solution is: Among them, α is expressed as: represents the second new electricity price baseline solution at the t+1th iteration in the i-th unit; Step 3.6.2: After aggregation, compare the fitness values ​​of the two electricity price baseline solutions before and after local mining to determine the update of the position; if the new solution is better than the original solution, update the position, otherwise do not perform the update. The mathematical expression of the position of the corresponding electricity price baseline solution is as follows: in, Indicates bringing in The magnitude of the direction vector is Indicates bringing in The magnitude of the direction vector; Step 3.7: Iterate optimization using gradient descent: Among them, Y old is the solution of the current electricity price baseline; Y new is the solution to the updated electricity price baseline; η is the learning rate, i.e., the step size, a positive constant that controls the magnitude of the update; is the objective function M in Y old The gradient at , expressed as a vector, points in the direction of the fastest growth of the function; Step 3.8: When the number of iterations reaches the maximum or the fitness value meets the requirements, exit the loop and output the optimal electricity price baseline, otherwise return to step 3.5 for a new round of iteration.

5. The intelligent power distribution device for industrial parks according to claim 4 is characterized in that: In step 3.4, in order to improve the search space of the algorithm, the formula for updating the size l of the triangular topological unit is as follows: Among them, t represents the current number of iterations; T represents the maximum number of iterations; l0 is the initial value of l; is a constant that controls the growth rate of l, and l decreases as the number of iterations increases.

6. The intelligent power distribution device for industrial parks according to claim 3 is characterized in that: The specific process of the intelligent control unit to achieve the allocation of power resources is as follows: When the difference between the total power generation at the power supply end of the industrial park transmission line and the power consumption of the industrial park is within [ΔP surplus ,+∞], the excess power resources are allocated to areas with insufficient power resources, or the electric energy is converted into other energy. When the power consumption peaks, other energy is converted into electric energy and sent to the power grid; when the difference between the total power generation at the power supply end of the transmission line connecting the industrial park and the power consumption of the industrial park is between [-∞,ΔP shortage ] When power is insufficient, electricity is transferred from other areas with abundant power resources, or additional equipment is added at the power supply end of the transmission line.

Images