Industrial load multi-round adjustment method based on safety evaluation

Abstract

The invention discloses an industrial load multi-round adjustment method based on safety evaluation, and belongs to the technical field of electrical engineering. According to the method, firstly, a safety electricity utilization risk index system is established by obtaining industrial user internal load key data, then risk assessment is carried out based on a self-learning weight model and a cloud model, and the self-learning weight model can automatically adjust the weight of each risk index according to historical data and real-time information. According to the method, the real-time requirement of a power grid is met by using a cloud model to reflect the actual importance of the power grid under the current condition, the cloud model comprehensively analyzes risk indexes by using the principles of fuzzy logic and probability theory to obtain a more accurate risk assessment result, finally, the load is regulated and controlled through a multi-round regulation method, and the load is dynamically regulated and controlled to adapt to the real-time requirement of the power grid. The method not only considers the power utilization safety of the industrial user, but also considers the economic operation of the power grid, and fully considers the problems of the high-energy-consumption industrial load in the regulation and control process, so as to maintain the safe and reliable operation of the power grid.

Description

Technical Field

[0001] The invention belongs to the technical field of electrical engineering, and in particular relates to a multi-round adjustment method for industrial loads based on safety assessment. Background Art

[0002] Under the background of the widespread application of new energy, the stability of the power grid faces unprecedented challenges. With the increasing proportion of renewable energy (such as solar energy, wind energy, etc.) in the power system, the uncertainty and volatility of the source-load of the power grid have increased significantly. This change not only affects the safety and reliability of the power system, but also puts higher requirements on power dispatching and load management. To meet this challenge, reasonable load regulation has become an important means to solve the source-load mismatch problem. In load-side regulation, industrial users have become an important source of adjustable resources due to their high electricity demand and diversified energy use. Compared with residential users, industrial users usually have more complex power equipment and production processes, which makes their loads flexible and can be adjusted accordingly according to the actual needs of the power grid. For example, during the peak load period of the power grid, some industrial users can reduce the production load in time, and during the period of low electricity demand, they can increase electricity consumption to improve production efficiency. However, the premise of load control is to ensure the safety and stability of the controlled load. When adjusting the load of industrial users, their production processes, equipment operation characteristics and safety standards must be fully considered to ensure that the regulation process will not affect production safety and the normal operation of equipment.

[0003] In order to achieve effective regulation of industrial users, it is crucial to implement multiple rounds of regulation. First, a comprehensive analysis of the electricity consumption characteristics of industrial users is required. This includes in-depth research on the dynamic change rules of loads, peak and valley load characteristics, and the relationship with the operating status of the power grid. By establishing data models and algorithms, it is possible to accurately predict load demand under different conditions, so as to formulate appropriate regulation strategies. Secondly, an efficient load regulation model is constructed to cope with grid load fluctuations or unstable new energy generation, ensuring that the electricity consumption behavior of industrial users can be adjusted in a timely manner, thereby maintaining the overall stability of the power grid.

[0004] Therefore, there is an urgent need in the prior art for a new technical solution to perform multiple rounds of adjustments on industrial users to ensure their safe use of electricity. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a multi-round adjustment method for industrial loads based on safety assessment, which is used to solve the technical problem in the prior art that it is difficult to effectively regulate the complex electricity consumption characteristics of industrial users, thereby affecting the safety of power grid operation.

[0006] The technical solution adopted by the present invention is to provide a multi-round adjustment method for industrial loads based on safety assessment, comprising:

[0007] Step 1: Obtain key internal load data of industrial users and establish a safe electricity use risk indicator system;

[0008] Step 2: Perform risk assessment based on the self-learning weight model and cloud model to obtain the optimal weight of the industrial user safety electricity use risk indicator system;

[0009] Step 3: Regulate the load through multiple rounds of regulation.

[0010] The step 1 includes:

[0011] Step 1.1: Establish indicator 1: load regulation rate;

[0012] In the process of load regulation, there are maximum and minimum limits for load regulation. A load regulation rate index is established to indicate the impact on the stable operation of the load in the process of load regulation, as shown in the following formula:

[0013]

[0014] Among them, P max , P min are the maximum and minimum limits of the load, P c is the actual regulated power of the load;

[0015] Step 1.2: Establish indicator 2: load adjustment time;

[0016] During the load regulation process, the safety of industrial users' production and operation will also be affected. According to the longest time that users can produce safely and stably during the regulation process, the total load regulation time index is constructed, as shown in the following formula:

[0017]

[0018] Among them, t c is the actual load regulation duration, t N is the maximum adjustable duration of the load;

[0019] Step 1.3: Establish indicator 3: actual adjustment percentage;

[0020] During the load regulation process, the production capacity of industrial users will be affected, and as the regulation amount increases, the ability of industrial users to produce safely will decrease. The actual regulation amount percentage index is constructed to evaluate the impact of user safety and stability on Sydney Harbour during the load regulation process, as shown in the following formula:

[0021]

[0022] Among them, P c is the actual voltage regulation power of the load, P N is the total power of the load;

[0023] Step 1.4: Establish indicator 4: load regulation frequency;

[0024] Set the maximum number of times that the adjustable load can be adjusted to ensure its safe operation when adjusted. Construct a load adjustment frequency index to characterize the risk of the impact of the adjustment number on the load operation status, as shown in the following formula:

[0025]

[0026] Among them, cs c is the actual load adjustment times, cs N is the maximum number of times the load can be adjusted;

[0027] Step 1.5: Establish indicator 5: total load remaining capacity limit;

[0028] Safety load is a type of load that ensures the smooth progress of industrial production. In industrial production, stable operation must be ensured and the total load remaining capacity index is constructed, as shown in the following formula:

[0029]

[0030] Among them, S ab is the total capacity of the safety load, S all is the remaining capacity of the total user load after control;

[0031] Step 1.6: Establish indicator 6: busbar overvoltage risk;

[0032] During the review and regulation, as the load is added and removed, the bus voltage where it is located will fluctuate to varying degrees. When the voltage fluctuation exceeds a certain range, the production of industrial users will be in danger. The bus overvoltage risk index is constructed to reflect the mutual influence relationship between the load regulation process and user safety, as shown in the following formula:

[0033]

[0034] Where ΔP represents the controllable active power of the load connected to the bus, ΔQ represents the reactive power of the load connected to the bus, R and X are the line resistance and reactance connected to the bus, and U N is the rated voltage of the busbar, ΔU c is the difference in bus voltage rise after the adjustable load is regulated, num is the number of overvoltage-sensitive loads among the remaining adjustable loads;

[0035] Step 1.7: Establish indicator 7: load correlation;

[0036] The production process of industrial users can be divided into continuous or discrete production. For continuous production, there is a close sequential relationship between loads. Once one of the equipment fails or operates abnormally, it will affect the associated loads. The load correlation index in the production process is constructed to characterize the risks caused by load interaction, as shown in the following formula:

[0037]

[0038] There are a total of n processes on the assembly line, where a is the degree of correlation between process s and process r, no correlation is considered 0, the degree to which process s is controlled but not to the point of completely stopping process r is recorded as 1, and the degree to which process s must run to meet the requirements of process r is recorded as 2. When a process is controlled, the corresponding row values ​​are added to show the degree of its influence on other loads;

[0039] Step 1.8: Establish indicator 8: uncertainty of renewable energy output;

[0040] Renewable energy generation is highly unstable and volatile, and it is difficult to predict the power of renewable energy generation. When there is a large fluctuation on the power supply side, it directly affects the stable operation of the load side. For this reason, a renewable energy output uncertainty index is constructed to represent the relationship between the stable operation of users and the fluctuation of renewable energy output, as shown in the following formula:

[0041] Z8=P t ·δ t+1

[0042] Among them, P t is the output power of renewable energy generation at that moment during load regulation, δ t+1 It is the probability that the new energy source cannot guarantee the original output at the next moment;

[0043] Step 1.9: Establish indicator 9: economic losses;

[0044] In the case of regulation, the load is still running, but the reduction in power leads to a decrease in productivity. In the case of failure risk, the load cannot run at this time, resulting in economic losses. The change in load size is approximated as the change in production scale, and a production shortage compensation index is constructed to characterize the economic risk brought about by abnormal load output, as shown in the following formula:

[0045]

[0046] Among them, P jN is the rated capacity of the jth load, P jc is the actual regulation power of the jth load, ρ is the contract production payment ratio, σ j is the flag to indicate whether the jth load is removed, σj =1 means the load is cut off, σ j =0 means the load is not removed, Cp j is the j-th load interruption loss cost, and a total of k loads are adjusted.

[0047] The step 2 includes:

[0048] Step 2.1: Establish a self-learning weight model based on the relative change rate of weights;

[0049] The self-learning weight model based on the relative change rate of weights includes an objective function based on the relative change rate of weights, an improvement of the subjective interval, the setting of constraint conditions and an optimization algorithm solution; specifically:

[0050] Step 2.1.1: Establish an objective function based on the relative rate of change of weights;

[0051] A single assignment method is used to combine the indicator weights of different assignment methods, as shown in the following formula:

[0052]

[0053] in, is the self-learning weight of the ith indicator, is the weight of the ith indicator of the dth weighted method, β id is the combination factor of the ith indicator of the dth weighting method, u is the total number of weighting methods, and n is the total number of indicators;

[0054] The relative change rate of the weights of different assignment methods among indicators is expressed by the following formula:

[0055]

[0056] Among them, k di is the relative change rate of the weights of the ith indicator and the i+1th indicator of the dth weighting method;

[0057] In order to effectively integrate the indicator weight information of different assignment methods, the objective function is constructed based on the relative change rate of the weight, as shown in the following formula:

[0058]

[0059] Step 2.1.2: Improve the subjective interval;

[0060] Assume that N indicators are weighted by M weighting methods and construct a weight matrix B, as shown in the following formula:

[0061]

[0062] in, is the weight of the ith indicator calculated by the dth weighting method;

[0063] The range of combined weights can be determined by the weight matrix B, The following formula can be used to obtain the subjective weight interval of N indicators:

[0064]

[0065] Step 2.1.3: Set constraints;

[0066] The indicators are divided into three categories: increase, decrease and uncertainty. The classification process is as follows:

[0067] Assume a dataset X with n indicators Σ ={X1,X2,…,X n} respectively correspond to the weights of different weighting methods W d The selected weight value is compared with the unselected weight value in the equation as shown below:

[0068]

[0069] Among them, b i is the category of the i-th indicator, {1, 2, 3} corresponds to increasing, decreasing and uncertainty respectively. Constraints can be constructed based on labels, as shown in the following formula:

[0070]

[0071] Step 2.1.4: Use the Grey Wolf Algorithm to solve;

[0072] The Gray Wolf Algorithm was proposed in 2014 as a heuristic algorithm, and its performance has been widely recognized in various engineering optimization problems. In the present invention, the Gray Wolf Algorithm is used to solve the optimal weight of the self-learning weight model;

[0073] Step 2.2: Establish the indicator risk level association matrix based on the cloud model;

[0074] The risk assessment index system for industrial users is quantified through the cloud model, the risk level is divided into four levels, and the specific index risk level range and cloud model parameters are determined;

[0075] Substitute the non-negative normalized indicator data set into the following formula to calculate the correlation R of each indicator under different risk levels: q =[r q1 ,r q2 ,r q3 ,r q4 ],(q=1,2,…,n);

[0076]

[0077] in, For E n The mathematical expectation of a normal random value, H e is the standard deviation, NORM is the normal random function;

[0078] Integrate the correlation matrix of each indicator to construct a risk assessment matrix, as shown in the following formula:

[0079]

[0080] Among them, n is the total number of indicators, r ij The correlation of the ith indicator belonging to the jth risk level;

[0081] Step 2.3: Combine the self-learning weight model and cloud model for risk assessment

[0082] The risk assessment process combining the self-learning weight model and the cloud model includes: determining the optimal combination weight vector, determining the risk assessment matrix, and determining the risk assessment classification according to the membership level; specifically:

[0083] Step 2.3.1: Determine the optimal weight vector;

[0084] Three different weighting methods are selected to calculate the combined weights, and the subjective weight W1 = [ω 11 ,ω 21 ,…,ω n1 ] T Calculate and use the entropy weight method and independence weight method to calculate the objective weight W2=[ω 12 ,ω 22 ,…,ω n2 ] T and W3=[ω 13 ,ω 23 ,…,ω n3 ] T Calculate separately;

[0085] The entropy weight method, hierarchical analysis method and independent weighting method are used for weighting, and the weight vector W of the optimal combination is obtained through the self-learning weight model. c , expressed by the following formula:

[0086]

[0087] in, is the optimal weight combination of the i-th indicator;

[0088] Step 2.3.2: Determine the risk assessment matrix and determine the risk assessment classification according to the membership level;

[0089] The optimal combination weight W c Risk assessment matrix R determined by cloud model Σ Combined, the risk assessment result of industrial users is expressed as follows:

[0090] Z s =W c ·R Σ ,(s=1,2,…,m)

[0091] Evaluation results Z of m industrial users s =[z s1 ,z s2 ,z s3 ,z s4 ], according to the maximum membership principle, Z s The maximum value in corresponds to the risk level of the industrial user, which is used as the risk assessment classification result of the industrial user.

[0092] In step 3, the industrial user load includes multiple loads, and the auxiliary production load is preferentially reduced during load control; if the load reduction is greater than the auxiliary production load, the excess is borne by the main production load;

[0093] Assuming that the total number of industrial users participating in multiple rounds of regulation is m, define the risk level reference limit FJX of controllable load capacity y =[fjx y1 ,fjx y2 ,fjx y3 ,fjx y4 ], (y=1,…,m) so as to quickly adjust the risk level by adjusting the controllable capacity of industrial users;

[0094] Among them, the maximum controllable capacity of the yth industrial user is used as the reference limit of the load control amount of the risk level, and 3 / 4, 1 / 2 and 1 / 4 of the maximum controllable capacity are used as the reference limit of the control capacity of risk levels 3, 2 and 1. y , combined with the reference limit value FJX of the control load capacity y , realizing multi-round regulation of industrial loads.

[0095] Through the above design scheme, the present invention can bring the following beneficial effects:

[0096] By constructing an industrial user safety electricity use risk indicator system, the safety risk level of industrial users' electricity use can be evaluated, and the load can be adjusted using the risk level reference limit of the controllable load capacity. Under safety constraints, multiple rounds of adjustments can be made to industrial users by adjusting auxiliary production loads and production loads to evaluate the risk of internal load adjustment of industrial users, achieve power balance, ensure the range of voltage fluctuations, and ensure safe industrial production. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 A flow chart of a multi-round industrial load regulation method based on safety assessment of the present invention;

[0098] Figure 2 It is a schematic diagram of an industrial user's safe electricity use risk situation indicator system of an industrial load multi-round adjustment method based on safety assessment of the present invention;

[0099] Figure 3 It is a flow chart of industrial user risk classification assessment of a multi-round industrial load adjustment method based on safety assessment of the present invention;

[0100] Figure 4 A specific voltage distribution diagram of nodes where high-energy-consuming industrial users are located during the third round of load regulation in Example 1 of a multi-round industrial load regulation method based on safety assessment of the present invention;

[0101] Figure 5 A specific voltage distribution diagram of nodes where high-energy-consuming industrial users are located during the second round of load regulation in Embodiment 2 of a multi-round industrial load regulation method based on safety assessment of the present invention;

[0102] Figure 6 This is a specific voltage distribution diagram of nodes where high-energy-consuming industrial users are located during the second round of load regulation in Example 3 of a multi-round industrial load regulation method based on safety assessment of the present invention. DETAILED DESCRIPTION

[0103] The present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0104] Embodiment 1:

[0105] In this embodiment, relying on the regional microgrid, high-energy-consuming users and new energy power generation systems are connected to different nodes. Taking the electrolytic aluminum industry as an example, according to research and data collection, the proportion of auxiliary production load to the total load is generally 4-6%. This embodiment takes 5%. In this embodiment, the power on the power supply side is 2680.5MW, and the power on the demand side is 2875MW. The power generation side cannot provide enough power to support the load. This embodiment adjusts the load three times. Figure 1 This is the industrial load multi-round regulation control flow chart considering safety constraints of the present invention. The information of 8 high-energy-consuming industrial users is shown in Table 1:

[0106] Table 18 Information on high energy consumption industrial users

[0107]

[0108] The multi-round regulation method of industrial load based on safety assessment includes:

[0109] Step 1: Calculate the load reduction capacity of eight high-energy-consuming industrial users using the optimal load reduction model;

[0110] Step 2: Obtain the reference limit of the controllable load according to the maximum controllable capacity;

[0111] Step 3: Deduct the reducible capacity of industrial users from the auxiliary production load to obtain the remaining reducible capacity;

[0112] Step 4: According to the reduced capacity of controllable industrial users, the corresponding controllable load reference limit and relevant parameters of industrial users can be obtained, and risk assessment is conducted on 8 industrial users. Figure 2 This is a schematic diagram of the risk situation indicator system for safe electricity use by industrial users based on the multi-round adjustment method of industrial loads based on safety assessment. Figure 3 It is a flow chart of risk classification assessment for industrial users based on the multi-round regulation method of industrial loads based on safety assessment;

[0113] Step 5: Set the controllable risk level limit and filter out industrial users that do not meet the controllable risk limit;

[0114] Step 6: Adjust the industrial load according to the industrial users that do not meet the conditions and the corresponding adjustable capacity under the controllable risk level, and repeat step 2 until the controllable risk threshold is reached;

[0115] Step 7: Perform power flow calculation on the power grid of the industrial user after load adjustment to determine whether the power provided by the power source is balanced with the power consumed by the load. If not, further adjust the industrial load (repeat steps 1 to 6) until the power provided by the power source meets the power consumed by the load.

[0116] Step 8: When the power of the power source and the load meets the balance requirements, calculate whether the bus node voltage of the industrial user remains within the normal range. If it exceeds the normal range, repeat steps 1 to 7 until the voltage of each industrial user node meets the requirements.

[0117] The specific implementation of the present invention is as follows:

[0118] Step 1 The specific steps are as follows:

[0119] According to the optimal load reduction model, the reduced capacity of all controllable industrial users is obtained. This embodiment includes 8 high-energy-consuming industrial users. The reduced capacity of industrial users is shown in Table 2:

[0120] Table 28 High Energy Consumption Industrial Users Reduced Capacity

[0121]

[0122] Step 2 The specific steps are as follows:

[0123] According to the maximum controllable capacity of 8 industrial users, the reference limit value FJX of controllable load is obtained y ,(y=1、2、…、8).

[0124] The total number of industrial users participating in multiple rounds of regulation is 8, and the risk level reference limit FJX of controllable load capacity is defined y =[fjx y1 ,fjx y2 ,fjx y3 ,fjx y4 ], (y=1,…,8) so as to quickly adjust the risk level by adjusting the controllable capacity of industrial users.

[0125] Among them, the maximum controllable capacity of the yth industrial user is used as the reference limit of the load control amount of the risk level, and 3 / 4, 1 / 2 and 1 / 4 of the maximum controllable capacity are used as the reference limit of the control capacity of risk levels 3, 2 and 1. y .

[0126] Step 3 The specific steps are as follows:

[0127] Deduct the reducible capacity of industrial users from the auxiliary production load, and obtain the remaining reducible capacity SXJ = [sxj1, sxj2, ..., sxj8]. y ≤0, there is no need to reduce the important load with adjustable capacity, and SXJ is taken y = 0, on the contrary, when it is necessary to reduce the adjustable important load, keep SXJ y to get the new SXJ.

[0128] Step 4 The specific steps are as follows:

[0129] According to the reduced capacity of controllable industrial users, the corresponding controllable load reference limit FJX and the relevant parameters of industrial users can be obtained, and the risk assessment index data set is obtained through power flow calculation. The risk assessment method combining the self-learning weight model and the cloud model is used to conduct risk assessment on all controllable industrial users, and the corresponding risk level set FD = [fd1, fd2, ..., fd8] is obtained. The weights of each indicator obtained by the self-learning weight model for 8 high-energy-consuming industrial users are shown in Table 3. The first load regulation risk assessment results of 8 industrial users obtained by using the cloud model are shown in Table 4.

[0130] The self-learning weight model is solved by the Grey Wolf Algorithm, and the Grey Wolf Algorithm model (GWO) includes social hierarchy, tracking, encirclement and attack on prey.

[0131] Specifically, the social hierarchy of the Gray Wolf Algorithm model includes:

[0132] The social hierarchy of wolves is mathematically modeled when designing GWO, and the best-fitting solution is considered as α. Therefore, the second and third best solutions are named β and δ respectively. The remaining candidate solutions are assumed to be ω. In the GWO algorithm, hunting (optimization) is guided by α, β and δ, and the wolf pack belonging to ω follows these three wolves.

[0133] Specifically, the encirclement prey of the gray wolf algorithm model includes:

[0134]

[0135]

[0136]

[0137]

[0138] Among them, t represents the current number of iterations, and is the coefficient vector, is the prey's position vector, is the position vector of the gray wolf, With the iteration, it linearly transforms from 2 to 1. and is a random vector in [0,1].

[0139] Specifically, the hunting of the Gray Wolf algorithm model includes:

[0140]

[0141]

[0142]

[0143] Specifically, the attack prey of the Gray Wolf algorithm model includes:

[0144] Gray wolves hunt by attacking their prey when it stops moving. In the process of building a prey attack model, they surround their prey. Will follow The value of decreases, The range of change is when When it is in the interval [-1,1], the next search position can be anywhere between the current gray wolf and the prey.

[0145] Specifically, the prey searched by the Gray Wolf Algorithm model includes:

[0146] The gray wolf mainly relies on the information of α, β and δ to find prey. By setting |A|>1, the search agent is kept away from the prey. This search method enables GWO to perform global search. Another search coefficient in the GWO algorithm is This coefficient prevents the algorithm from falling into a local optimum.

[0147] Table 3 Weight values ​​of each indicator

[0148]

[0149] Table 4 The first load regulation risk assessment

[0150]

[0151] Step 5 The specific steps are as follows:

[0152] Set the controllable risk level limit θ, compare the elements in the risk level set FD with 1, and filter out industrial users FD that do not meet the controllable risk limit α ,From Table 4, we can see that users 1 and 2 do not meet the requirements.

[0153] Step 6 The specific steps are as follows:

[0154] According to FD α Select industrial users and adjust the reduced capacity to the same value as fjx in combination with the controllable risk level limit θ uθ The load control quantity reference limit θ corresponding to the risk level. Repeat step 2; if the risk level is still not reached, further reduce the capacity to The corresponding load control reference limit is repeated in step 2 until the controllable risk threshold is reached. The results of industrial users whose load adjustment does not reach the controllable risk threshold in the first time reducing the adjustment and re-assessing the risk are shown in Table 5.

[0155] Table 5 Results of reducing adjustments and re-assessing risks for industrial users that did not reach the controllable risk threshold

[0156]

[0157] Step 7 The specific steps are as follows:

[0158] The power flow of the power grid of the industrial users after load reduction is calculated, and the power provided by the power source is compared with the power consumed by the load. If the power generation side cannot meet the demand of the power supply side, the load is further adjusted (repeat steps 1-6) until the power generation side meets the demand of the load side. After the first load adjustment of the 8 high-energy-consuming industrial users, the power on the power source side was 2680.5MW and the power on the load side was 2756.956MW. At this time, the source-load power is unbalanced, and the second load adjustment is required. The reduction in user capacity, the weight of each indicator, the risk assessment level, and the results of the reduction adjustment and re-risk assessment of industrial users who have not reached the controllable risk limit in the second load adjustment are shown in Tables 6, 7, 8, and 9, respectively:

[0159] Table 68 High Energy Consumption Industrial Users Reduce Capacity

[0160]

[0161]

[0162] Table 7 Weight values ​​of each indicator

[0163]

[0164] Table 8 Second load regulation risk assessment

[0165]

[0166] Table 9 Results of reduction adjustment and re-assessment of risk for industrial users that did not reach the controllable risk threshold

[0167]

[0168] After the second load adjustment, the power of the 8 high-energy-consuming industrial users is 2705MW on the power side and 2690.789MW on the load side. At this time, the source-load power is balanced from a numerical point of view, and the voltage of the node where the high-energy-consuming industrial users are located also meets the requirements. However, considering the line loss, the power side needs to provide 2717.170MW of power, and the power on the power side and the load side is still unbalanced, so the third load adjustment is required. The reduction in user capacity and the weight of each indicator in the third load adjustment are shown in Tables 10 and 11.

[0169] Table 108 High Energy Consumption Industrial Users Reduce Capacity

[0170]

[0171] Table 11 Weight values ​​of each indicator

[0172]

[0173] After the third load adjustment, the power on the power supply side is 2680.5MW, the power on the load side is 2675.453MW, the line loss is 30.068MW, and the source-load power meets the balance requirements.

[0174] In summary, the multi-round adjustment method of industrial load based on safety assessment can perform multiple rounds of adjustments on the safe electricity consumption of industrial users according to the risk assessment results of the controlled industrial load users, so as to maintain the safe and reliable operation of the power grid.

[0175] Step 8 The specific steps are as follows:

[0176] When the source-load power meets the balance requirement, it is determined whether the voltage of each node is within the limit. If not, repeat steps 1 to 7 until the voltage of each node meets the requirement and the load control is terminated. The specific distribution of the node voltages of the eight high-energy-consuming industrial users is as follows: Figure 4 shown.

[0177] Example 2

[0178] In this embodiment, a data set of six electrolytic magnesium industries in a certain area of ​​Shandong is used to perform multiple rounds of industrial load adjustment. In this embodiment, the power on the power supply side is 3865MW, and the power on the demand side is 4075MW. The power generation side cannot provide enough power to support the load. In this embodiment, two load adjustments are performed on the six high-energy-consuming industrial users. The information of the six high-energy-consuming industrial users is shown in Table 12.

[0179] Table 1 Information on 26 high energy consuming industrial users

[0180]

[0181] The industrial load is regulated by using a multi-round industrial load regulation method based on safety assessment, the regulation method comprising the following steps:

[0182] Step 1: Obtain the load information of six high-energy-consuming industrial users and obtain the reduction amount of all industrial users according to the optimal load reduction model;

[0183] Step 2: Obtain a controllable load reference value based on the maximum controllable load, deduct the reducible capacity from the auxiliary production load, and determine whether it is necessary to reduce the important load;

[0184] Step 3: Use the risk assessment method combining the self-learning weight model and the cloud model to conduct risk assessment on six high-energy-consuming industrial users;

[0185] Step 4: Set controllable risk level limits, screen out industrial users that do not meet the controllable risk limits, adjust the industrial users, and re-evaluate the risks;

[0186] Step 5: Perform power flow analysis on the power grid where the industrial users are located after load adjustment to determine whether the source-load power is balanced; when the source-load power meets the requirements, determine whether the voltage distribution of the six industrial users is within the limit.

[0187] All load information data in the present invention are provided by local industrial users. The specific implementation methods of the present invention are as follows:

[0188] Step 1 The specific steps are as follows:

[0189] According to the load reduction model, the reduced capacity of all controllable industrial users is obtained, and the user reduced capacity is shown in Table 13.

[0190] Table 136 High Energy Consumption Industrial Users Reduced Capacity

[0191]

[0192] Step 2 The specific steps are as follows:

[0193] Step 2.1: Obtain a reference limit value of the controllable load according to the maximum controllable capacity;

[0194] Determine the risk level reference limit FJX of controllable load capacity y =[fjx y1 ,fjx y2 ,fjx y3 ,fjx y4 ], (y=1,…,6) so as to quickly adjust the risk level by adjusting the controllable capacity of industrial users.

[0195] Step 2.2: Subtract the load that can be reduced;

[0196] Deduct the reducible capacity of industrial users from the auxiliary production load, and obtain the remaining reducible capacity SXJ = [sxj1, sxj2, ..., sxj8]. y ≤0, there is no need to reduce the important load with adjustable capacity, and SXJ is taken y = 0, on the contrary, when it is necessary to reduce the adjustable important load, keep SXJ y to get the new SXJ.

[0197] Step 3 The specific steps are as follows:

[0198] A risk assessment method combining the self-learning weight model and the cloud model was used to conduct risk assessment on six high-energy-consuming industrial users. The weight indicators are shown in Table 14, and the results of the first load regulation risk assessment are shown in Table 15.

[0199] Table 14 Weight values ​​of each indicator

[0200]

[0201] Table 15 The first load regulation risk assessment

[0202]

[0203] Step 4 The specific steps are as follows:

[0204] Step 4.1: Screening out industrial users that do not meet the requirements;

[0205] Set the controllable risk level limit to screen out industrial users that do not meet the controllable risk limit. As shown in Table 15, users 1 and 2 do not meet the requirements.

[0206] Step 4.2: Adjust industrial users that do not meet the requirements;

[0207] According to the controllable risk level limit, the capacity reduction of users 1 and 2 is adjusted to the corresponding load control reference limit, and step 4.1 is repeated to determine whether the controllable risk limit has been reached. The results of the first load adjustment of users 1 and 2 and the re-risk assessment are shown in Table 16:

[0208] Table 16 Results of industrial users reducing adjustments and re-assessing risks

[0209]

[0210] Step 5 The specific steps are as follows:

[0211] Step 5.1: Determine whether the source-load power meets the requirements;

[0212] Conduct a flow analysis on the power grid where the industrial users are located after load adjustment, update the reference threshold of each user's maximum controllable capacity, auxiliary production load capacity and load control quantity, and judge whether the source-load power is balanced. After the first load adjustment of the six high-energy-consuming industrial users, the power on the power supply side is 3865MW, and the power on the demand side is 3976.638MW, which does not meet the requirements and needs to be adjusted for the second time. The results of the second load adjustment of industrial users to reduce adjustments and re-evaluate risks are shown in Table 17:

[0213] Table 17 Results of industrial users reducing adjustments and re-assessing risks

[0214]

[0215] After the second load adjustment, the power on the power supply side is 3865MW, the power on the load side is 3698.768MW, the line loss is 64.264MW, and the source-load power meets the balance requirements.

[0216] Step 5.2: Determine whether the voltage meets the requirements;

[0217] The source-load power meets the balance requirement. Analyze whether the bus voltage of the six industrial users is within the limit. The specific voltage distribution is as follows: Figure 5 As shown, the voltage meets the requirements.

[0218] In summary, it is shown that the multi-round adjustment method of industrial load based on safety assessment can ensure the safe production of users by adjusting their electricity consumption.

[0219] Example 3

[0220] In this embodiment, a dataset of six cement industries in Liaoning is used to control users through a multi-round industrial load adjustment method based on safety assessment. In this embodiment, the power on the power supply side is 2695MW and the power on the demand side is 3065MW. This embodiment achieves source-load power balance through two load adjustments. The information of the six high-energy-consuming industrial users is shown in Table 18.

[0221] Table 186 Information on high energy consumption industrial users

[0222]

[0223] The industrial load multi-round adjustment method based on safety assessment is selected to control the industrial load. The specific steps of the control method are as follows:

[0224] Step 1: Obtain the reduction capacity of all controllable industrial loads according to the optimal load reduction model, and obtain the reference value of the controllable load according to the maximum controllable capacity;

[0225] Step 2: Use the risk assessment method that combines the self-learning weight model and the cloud model to conduct risk assessment on six high-energy-consuming industrial users.

[0226] Step 3: Set controllable risk limits to screen out users who do not meet the requirements. After screening, user 2 and user 3 do not meet the requirements. Users 3 and 2 are controlled and risk assessment is re-performed. The results are shown in Table 19.

[0227] Table 19 Results of industrial users reducing adjustments and re-assessing risks

[0228]

[0229] Step 4: Perform power flow analysis based on the adjusted load to determine whether the source-load power is balanced;

[0230] After the first load adjustment, the power on the power supply side is 2695MW and the power on the load side is 2806MW, which does not meet the requirements. Steps 1 to 4 are repeated. The results of the second user adjustment and re-risk assessment are shown in Table 20.

[0231] Table 20 Results of industrial users reducing adjustments and re-assessing risks

[0232]

[0233] After the second load adjustment, the power on the power generation side is 2695MW, the power on the load side is 2624.894MW, and the line loss is 20.683MW, which meets the source-load power balance requirement. The specific voltage distribution of the six industrial users is as follows: Figure 6 As shown, it is stable within the limit.

[0234] In summary, the multi-round regulation method of industrial loads based on safety assessment can not only ensure that industrial users are at a controllable safety risk level and maintain their safety, but also meet the requirements of safe and reliable operation of the power grid itself.

[0235] The implementation methods of the present invention are not limited to the above-mentioned embodiments, and any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A multi-round adjustment method for industrial loads based on safety assessment, characterized in that: include: Step 1: Obtain key internal load data of industrial users and establish a safe electricity use risk indicator system; Step 2: Perform risk assessment based on the self-learning weight model and cloud model to obtain the optimal weight of the industrial user safety electricity use risk indicator system; Step 3: Regulate the load through multiple rounds of regulation.

2. The method for multi-round industrial load regulation based on safety assessment according to claim 1 is characterized in that: The step 1 includes: Step 1.1: Establish indicator 1: load regulation rate; In the process of load regulation, there are maximum and minimum limits for load regulation. A load regulation rate index is established to indicate the impact on the stable operation of the load in the process of load regulation, as shown in the following formula: Among them, P max , P min are the maximum and minimum limits of the load, P c is the actual regulated power of the load; Step 1.2: Establish indicator 2: load adjustment time; During the load regulation process, the safety of industrial users' production and operation will also be affected. According to the longest time that users can produce safely and stably during the regulation process, the total load regulation time index is constructed, as shown in the following formula: Among them, t c is the actual load regulation duration, t N is the maximum adjustable duration of the load; Step 1.3: Establish indicator 3: actual adjustment percentage; During the load regulation process, the production capacity of industrial users will be affected, and as the regulation amount increases, the ability of industrial users to produce safely will decrease. The actual regulation amount percentage index is constructed to evaluate the impact of user safety and stability on Sydney Harbour during the load regulation process, as shown in the following formula: Among them, P c is the actual voltage regulation power of the load, P N is the total power of the load; Step 1.4: Establish indicator 4: load regulation frequency; Set the maximum number of times that the adjustable load can be adjusted to ensure its safe operation when adjusted. Construct a load adjustment frequency index to characterize the risk of the impact of the adjustment number on the load operation status, as shown in the following formula: Among them, cs c is the actual number of load adjustments, cs N is the maximum number of times the load can be adjusted; Step 1.5: Establish indicator 5: total load remaining capacity limit; Safety load is a type of load that ensures the smooth progress of industrial production. In industrial production, stable operation must be ensured and the total load remaining capacity index is constructed, as shown in the following formula: Among them, S ab is the total capacity of the safety load, S all is the remaining capacity of the total user load after control; Step 1.6: Establish indicator 6: busbar overvoltage risk; During the review and regulation, as the load is added and removed, the bus voltage where it is located will fluctuate to varying degrees. When the voltage fluctuation exceeds a certain range, the production of industrial users will be in danger. The bus overvoltage risk index is constructed to reflect the mutual influence relationship between the load regulation process and user safety, as shown in the following formula: Where ΔP represents the controllable active power of the load connected to the bus, ΔQ represents the reactive power of the load connected to the bus, R and X are the line resistance and reactance connected to the bus, and U N is the rated voltage of the busbar, ΔU c is the difference in bus voltage rise after the adjustable load is regulated, num is the number of overvoltage-sensitive loads among the remaining adjustable loads; Step 1.7: Establish indicator 7: load correlation; The production process of industrial users can be divided into continuous or discrete production. For continuous production, there is a close sequential relationship between loads. Once one of the equipment fails or operates abnormally, it will affect the associated loads. The load correlation index in the production process is constructed to characterize the risks caused by load interaction, as shown in the following formula: There are a total of n processes on the assembly line, where a is the degree of correlation between process s and process r, no correlation is considered 0, the degree to which process s is controlled but not to the point of completely stopping process r is recorded as 1, and the degree to which process s must run to meet the requirements of process r is recorded as 2. When a process is controlled, the corresponding row values ​​are added to show the degree of its influence on other loads; Step 1.8: Establish indicator 8: uncertainty of renewable energy output; Renewable energy generation is highly unstable and volatile, and it is difficult to predict the power of renewable energy generation. When there is a large fluctuation on the power supply side, it directly affects the stable operation of the load side. For this reason, a renewable energy output uncertainty index is constructed to represent the relationship between the stable operation of users and the fluctuation of renewable energy output, as shown in the following formula: Z8=P t ·δ t+1 Among them, P t is the output power of renewable energy generation at that moment during load regulation, δ t+1 It is the probability that the new energy source cannot guarantee the original output at the next moment; Step 1.9: Establish indicator 9: economic losses; In the case of regulation, the load is still running, but the reduction in power leads to a decrease in productivity. In the case of failure risk, the load cannot run at this time, resulting in economic losses. The change in load size is approximated as the change in production scale, and a production shortage compensation index is constructed to characterize the economic risk brought about by abnormal load output, as shown in the following formula: Among them, P jN is the rated capacity of the jth load, P jc is the actual regulation power of the jth load, ρ is the contract production payment ratio, σ j is the flag to indicate whether the jth load is removed, σ j =1 means the load is cut off, σ j =0 means the load is not removed, Cp j is the j-th load interruption loss cost, and a total of k loads are adjusted.

3. The method for multi-round regulation of industrial loads based on safety assessment according to claim 1, characterized in that: The step 2 includes: Step 2.1: Establish a self-learning weight model based on the relative change rate of weights; The self-learning weight model based on the relative change rate of weights includes an objective function based on the relative change rate of weights, an improvement of the subjective interval, the setting of constraint conditions and an optimization algorithm solution; specifically: Step 2.1.1: Establish an objective function based on the relative rate of change of weights; A single assignment method is used to combine the indicator weights of different assignment methods, as shown in the following formula: in, is the self-learning weight of the ith indicator, is the weight of the ith indicator of the dth weighted method, β id is the combination factor of the ith indicator of the dth weighting method, u is the total number of weighting methods, and n is the total number of indicators; The relative change rate of the weights of different assignment methods among indicators is expressed by the following formula: Among them, k di is the relative change rate of the weights of the ith indicator and the i+1th indicator of the dth weighting method; In order to effectively integrate the indicator weight information of different assignment methods, the objective function is constructed based on the relative change rate of the weight, as shown in the following formula: Step 2.1.2: Improve the subjective interval; Assume that N indicators are weighted by M weighting methods and construct a weight matrix B, as shown in the following formula: in, is the weight of the ith indicator calculated by the dth weighting method; The range of combined weights can be determined by the weight matrix B, The following formula can be used to obtain the subjective weight interval of N indicators: Step 2.1.3: Set constraints; The indicators are divided into three categories: increase, decrease and uncertainty. The classification process is as follows: Assume a dataset X with n indicators Σ ={X1,X2,…,X n } respectively correspond to the weights of different weighting methods W d The selected weight value is compared with the unselected weight value in the equation as shown below: Among them, b i is the category of the i-th indicator, {1, 2, 3} corresponds to increasing, decreasing and uncertainty respectively. Constraints can be constructed based on labels, as shown in the following formula: Step 2.1.4: Use the Grey Wolf Algorithm to solve; The Gray Wolf Algorithm was proposed in 2014 as a heuristic algorithm, and its performance has been widely recognized in various engineering optimization problems. In the present invention, the Gray Wolf Algorithm is used to solve the optimal weight of the self-learning weight model; Step 2.2: Establish the indicator risk level association matrix based on the cloud model; The risk assessment index system for industrial users is quantified through the cloud model, the risk level is divided into four levels, and the specific index risk level range and cloud model parameters are determined; Substitute the non-negative normalized indicator data set into the following formula to calculate the correlation R of each indicator under different risk levels: q =[r q1 ,r q2 ,r q3 ,r q4 ],(q=1,2,…,n); in, For E n The mathematical expectation of a normal random value, H e is the standard deviation, NORM is the normal random function; Integrate the correlation matrix of each indicator to construct a risk assessment matrix, as shown in the following formula: Among them, n is the total number of indicators, r ij The correlation of the ith indicator belonging to the jth risk level; Step 2.3: Combine the self-learning weight model and cloud model for risk assessment The risk assessment process combining the self-learning weight model and the cloud model includes: determining the optimal combination weight vector, determining the risk assessment matrix, and determining the risk assessment classification according to the membership level; specifically: Step 2.3.1: Determine the optimal weight vector; Three different weighting methods are selected to calculate the combined weights, and the subjective weight W1 = [ω 11 ,ω 21 ,…,ω n1 ] T Calculate and use the entropy weight method and independence weight method to calculate the objective weight W2=[ω 12 ,ω 22 ,…,ω n2 ] T and W3=[ω 13 ,ω 23 ,…,ω n3 ] T Calculate separately; The entropy weight method, hierarchical analysis method and independent weighting method are used for weighting, and the weight vector W of the optimal combination is obtained through the self-learning weight model. c , expressed by the following formula: in, is the optimal weight combination of the i-th indicator; Step 2.3.2: Determine the risk assessment matrix and determine the risk assessment classification according to the membership level; The optimal combination weight W c Risk assessment matrix R determined by cloud model Σ Combined, the risk assessment result of industrial users is expressed as follows: Z s =W c ·R Σ ,(s=1,2,…,m) Evaluation results Z of m industrial users s =[z s1 ,z s2 ,z s3 ,z s4 ], according to the maximum membership principle, Z s The maximum value in corresponds to the risk level of the industrial user, which is used as the risk assessment classification result of the industrial user.

4. The method for multi-round industrial load regulation based on safety assessment according to claim 1 is characterized in that: In step 3, the industrial user load includes multiple loads, and the auxiliary production load is preferentially reduced during load control; if the load reduction is greater than the auxiliary production load, the excess is borne by the main production load; Assuming that the total number of industrial users participating in multiple rounds of regulation is m, define the risk level reference limit FJX of controllable load capacity y =[fjx y1 ,fjx y2 ,fjx y3 ,fjx y4 ], (y=1,…,m) so as to quickly adjust the risk level by adjusting the controllable capacity of industrial users; Among them, the maximum controllable capacity of the yth industrial user is used as the reference limit of the load control amount of the risk level, and 3 / 4, 1 / 2 and 1 / 4 of the maximum controllable capacity are used as the reference limit of the control capacity of risk levels 3, 2 and 1. y , combined with the reference limit value FJX of the control load capacity y , realizing multi-round regulation of industrial loads.

Images