Hot start distributed acceleration out-clearing method, device and medium for power demand side flexibility sharing transaction

Abstract

The application discloses a hot start distributed acceleration out-clearing method, equipment and medium for power demand side flexibility sharing transaction, relates to the field of power demand side, constructs a first out-clearing model and a second out-clearing model for power demand side flexibility sharing, solves the second out-clearing model, and outputs a current iteration initial value; calculates a flexibility demand remainder, sends the current iteration initial value and the flexibility demand remainder to a client, updates a flexibility supply average value vector and the flexibility demand remainder based on a flexibility supply variable updated and sent by the client; updates the Lagrange multiplier by using the updated flexibility demand remainder; judges whether the updated Lagrange multiplier and the flexibility supply average value vector meet a convergence condition, if not, returns to the client to update the flexibility supply variable, and if yes, generates an out-clearing result, protects the privacy of the client, realizes efficient distributed out-clearing for the flexibility sharing first out-clearing model, and is suitable for large-scale users participating in flexibility sharing.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power demand side, in particular to a hot start distributed acceleration out-clearing method, equipment and medium for power demand side flexibility sharing transaction. BACKGROUND

[0002] The power demand side flexibility sharing transaction can significantly improve the economy, reliability and adequacy of demand response, and the flexibility sharing out-clearing solution is the basis for realizing the flexibility sharing transaction. The current out-clearing and solution methods include: centralized solution, such as using interior point method or directly calling commercial solver, but in actual application, it is not conducive to the protection of user privacy, which may lead to the leakage of user privacy, and the calculation burden is heavy, the centralized solution will bring huge calculation burden to the organizer, and in the large-scale scene, the application is limited, and the constraint conditions reported by the user may exist in the non-convex situation, which will lead to the difficulty of directly solving the model; directly introducing ADMM (Alternating Direction Method of Multipliers) algorithm to solve the market out-clearing model in a distributed manner, which still faces the problems of high iteration number and long convergence time in the large-scale user participation scene; using reinforcement learning to accelerate out-clearing, which may lead to insufficient privacy protection and privacy leakage risk, and the black box characteristics of the reinforcement learning model may lead to poor controllability and interpretability of the results, especially in the complex market out-clearing, which may lead to out-clearing result compliance and credibility problems, and the performance of the model depends on the quality and coverage of the training data, if the training data is insufficient or incomplete, it may lead to result deviation, and even cannot converge.

[0003] From the above, how to solve the problems of user privacy leakage, heavy calculation burden and slow convergence, adapt to large-scale user participation in flexibility sharing, realize efficient distributed out-clearing solution, and improve the accuracy and efficiency of sharing transaction is a problem to be solved in the field. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a hot start distributed acceleration out-clearing method, equipment and medium for power demand side flexibility sharing transaction, which can solve the problems of user privacy leakage, heavy calculation burden and slow convergence, adapt to large-scale user participation in flexibility sharing, realize efficient distributed out-clearing solution, and improve the accuracy and efficiency of sharing transaction. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses a hot start distributed acceleration out-clearing method for power demand side flexibility sharing transaction, comprising:

[0006] constructing a first out-clearing model for power demand side flexibility sharing;

[0007] constructing a power demand side flexibility sharing second dispatching model based on the power demand side flexibility sharing first dispatching model, solving the power demand side flexibility sharing second dispatching model to output a current iteration initial value; the current iteration initial value comprises a Lagrange multiplier and a flexibility supply average value vector;

[0008] calculating a flexibility demand surplus by using the current iteration initial value, and sending the current iteration initial value and the flexibility demand surplus to a client, so that the client calculates a flexibility supply variable based on the current iteration initial value and the flexibility demand surplus on the basis of a user individual feasible region constraint condition;

[0009] updating the flexibility supply average value vector and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client to obtain an updated flexibility demand surplus;

[0010] updating the Lagrange multiplier by using the updated flexibility demand surplus to obtain an updated Lagrange multiplier;

[0011] judging whether the updated Lagrange multiplier and the flexibility supply average value vector meet a preset convergence condition, if the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition, generating a dispatching result based on the flexibility supply variable, and if the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, jumping to the process of calculating the flexibility demand surplus by using the current iteration initial value, and returning the current Lagrange multiplier, the flexibility demand surplus and the flexibility supply average value to the client, so that the client updates the flexibility supply variable;

[0012] implementing a sharing transaction of power demand side distributed dispatching by using the dispatching result.

[0013] Optionally, the constructing the power demand side flexibility sharing first dispatching model comprises:

[0014] constructing the power demand side flexibility sharing first dispatching model based on an objective function, a coupling constraint condition and a non-coupling constraint condition; the coupling constraint condition comprises a supply-demand balance constraint condition; and the non-coupling constraint condition comprises an energy regulation constraint condition, a supply vector direction constraint condition and a user individual feasible region constraint condition;

[0015] Correspondingly, the constructing the power demand side flexibility sharing second dispatching model comprises:

[0016] constructing the power demand side flexibility sharing second dispatching model based on the objective function, the supply-demand balance constraint condition, the energy regulation constraint condition and the supply vector direction constraint condition.

[0017] Optionally, the first clearing model is:

[0018] ;

[0019] wherein, is a flexibility supply vector of the user, is an adjustment cost coefficient of the user in a time period, is an adjustment cost of the user further adjusting power to provide flexibility, is a total demand of the flexibility sharing, is a Lagrange multiplier of a supply-demand balance constraint condition, is an individual operation domain of the user , is a Hadamard product operation. Optionally, the process of calculating the Lagrange multiplier comprises:

[0020] constructing an augmented Lagrange function based on the objective function and the coupling constraint condition;

[0021] calculating the Lagrange multiplier by using the augmented Lagrange function.

[0022]

[0023] Optionally, the augmented Lagrange function is:

[0024] ;

[0025] wherein, is a flexibility supply vector of the user, is a Lagrange multiplier corresponding to a supply-demand balance constraint condition, , is an adjustment cost of the user further adjusting power to provide flexibility, is a total demand of the flexibility sharing, is a penalty factor, .

[0026] Optionally, the calculating the flexibility demand surplus by using the current iteration initial value comprises:

[0027] calculating the flexibility demand surplus by using a flexibility demand surplus calculation formula on the current iteration initial value to obtain the flexibility demand surplus;

[0028] the flexibility demand surplus calculation formula is: ​​

[0029] ;

[0030] wherein, is the flexibility demand residual after the kth iteration, is the flexibility supply variable.

[0031] Optionally, the judging whether the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition comprises:

[0032] judging whether a norm square of a primal residual of the updated Lagrange multiplier is not greater than a preset primal residual convergence threshold;

[0033] and judging whether a norm square of a dual residual of the flexibility supply average value vector is not greater than a preset dual residual convergence threshold.

[0034] Optionally, if the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, the method further comprises:

[0035] updating a current iteration number, returning the current Lagrange multiplier, the flexibility demand residual and the flexibility supply average value to the client, so that the client calculates the updated flexibility supply variable based on the returned current Lagrange multiplier, the flexibility demand residual and the flexibility supply average value;

[0036] jumping to the process of updating the flexibility supply average value vector and the flexibility demand residual based on the flexibility supply variable updated and sent by the client, until the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition.

[0037] In a second aspect, the present application discloses an electronic device, comprising:

[0038] a memory configured to save a computer program;

[0039] a processor configured to execute the computer program to implement the power demand side flexibility sharing transaction warm start distributed acceleration clearing method.

[0040] In a third aspect, the present application discloses a computer storage medium configured to save a computer program; wherein the computer program is executed by a processor to implement the steps of the power demand side flexibility sharing transaction warm start distributed acceleration clearing method disclosed above.

[0041] It can be seen that the application provides a hot start distributed acceleration clearing method for power demand side flexibility sharing transaction, which comprises the following steps: constructing a power demand side flexibility sharing first clearing model; constructing a power demand side flexibility sharing second clearing model based on the power demand side flexibility sharing first clearing model, solving the power demand side flexibility sharing second clearing model to output a current iteration initial value; the current iteration initial value comprises a Lagrange multiplier and a flexibility supply average value vector; calculating a flexibility demand surplus by using the current iteration initial value, sending the current iteration initial value and the flexibility demand surplus to a client, so that the client calculates a flexibility supply variable based on the current iteration initial value and the flexibility demand surplus on the basis of a user individual feasible region constraint condition; updating the flexibility supply average value vector and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client to obtain an updated flexibility demand surplus; updating the Lagrange multiplier by using the updated flexibility demand surplus to obtain an updated Lagrange multiplier; judging whether the updated Lagrange multiplier and the flexibility supply average value vector meet a preset convergence condition, if the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition, generating a clearing result based on the flexibility supply variable; if the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, jumping to the process of calculating the flexibility demand surplus by using the current iteration initial value, returning the current Lagrange multiplier, the flexibility demand surplus and the flexibility supply average value to the client, so that the client updates the flexibility supply variable; and realizing a sharing transaction of power demand side distributed clearing by using the clearing result.The application aims at the problems of centralized clearing being not conducive to user privacy protection, the heavy computing burden of organizers, the large number of iterations and slow convergence of distributed solving based on an alternating direction multiplier algorithm, and the difficulty of existing accelerated clearing methods in being applied to flexible sharing distributed clearing, constructs a power demand side flexible sharing first clearing model, constructs a power demand side flexible sharing second clearing model based on the power demand side flexible sharing first clearing model, solves the power demand side flexible sharing second clearing model to output a current iteration initial value, the current iteration initial value includes a Lagrange multiplier and a flexible supply average value vector, which can reduce the redundant steps of iteration solving from zero of the alternating direction multiplier algorithm, calculates a flexible demand surplus by using the current iteration initial value, and sends the current iteration initial value and the flexible demand surplus to a client, so that the client calculates a flexible supply variable based on the current iteration initial value and the flexible demand surplus on the basis of a user individual feasible region constraint condition, updates the flexible supply average value vector and the flexible demand surplus based on the flexible supply variable, obtains an updated flexible demand surplus, updates the Lagrange multiplier by using the updated flexible demand surplus, obtains an updated Lagrange multiplier, judges whether the updated Lagrange multiplier and the flexible supply average value vector meet a preset convergence condition, if the preset convergence condition is met, generates a clearing result based on the flexible supply variable, realizes distributed clearing solving, adapts to large-scale user participation in flexible sharing, and improves the accuracy and efficiency of shared transactions. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only constitute the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of the provided drawings.

[0043] Figure 1 A flow chart of a power demand side flexible sharing transaction hot start distributed accelerated clearing method disclosed by the present application;

[0044] Figure 2 A flow chart of a flexible sharing mechanism distributed clearing based on an alternating direction multiplier algorithm disclosed by the present application;

[0045] Figure 3 A graph of purchase electricity curves of four users in the case of no DR according to their own electricity consumption habits or maximum economic benefits;

[0046] Figure 4 A graph of iteration convergence process of actual values and logarithmic values with base 10 and original residual errors, wherein Figure 4(a) is an iterative convergence process diagram of actual value and original residual error disclosed by the application, Figure 4 (b) is an iterative convergence process diagram of base-10 logarithmic value and original residual error disclosed by the application,

[0047] Figure 5 is an iterative convergence process diagram of actual value and dual residual error disclosed by the application, Figure 5 (a) is an iterative convergence process diagram of actual value and dual residual error disclosed by the application, Figure 5 (b) is an iterative convergence process diagram of base-10 logarithmic value and dual residual error disclosed by the application,

[0048] Figure 6 is a flexible sharing cost comparison diagram of centralized and distributed clearing disclosed by the application,

[0049] Figure 7 is an electronic device structure diagram provided by the application. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0051] The power demand side flexibility sharing transaction can significantly improve the economy, reliability and adequacy of demand response, and the flexibility sharing clearing solution is the basis for realizing the flexibility sharing transaction. The current clearing and solving methods include: centralized solving, such as using the interior point method or directly calling a commercial solver, but in actual application, it is not conducive to the protection of user privacy, which may lead to the leakage of user privacy, the calculation burden is heavy, the centralized solving will bring a huge calculation burden to the organizer, and in a large-scale scenario, the application is restricted, and the constraint conditions reported by the user may be non-convex, which will lead to the difficulty of directly solving the model; directly introducing ADMM algorithm for distributed solving of market clearing model, still facing the problems of high iteration number and long convergence time in the scenario of large-scale user participation; using reinforcement learning to accelerate clearing, which may lead to insufficient privacy protection and privacy leakage risk, and the black box characteristics of the reinforcement learning model may lead to poor controllability and interpretability of the results, especially in the complex market clearing, which may lead to clearing result compliance and credibility problems, and the performance of the model depends on the quality and coverage of the training data, if the training data is insufficient or incomplete, it may lead to result deviation, and even cannot converge. As can be seen from the above, how to solve the problems of user privacy leakage, heavy calculation burden and slow convergence, adapt to large-scale user participation in flexibility sharing, realize efficient distributed clearing solution, and improve the accuracy and efficiency of sharing transaction are problems to be solved in the field.

[0052] Referring to Figure 1 The embodiment of the present application discloses a hot start distributed acceleration clearing method for power demand side flexibility sharing transaction, which can specifically include:

[0053] Step S11: constructing a first clearing model for power demand side flexibility sharing.

[0054] In this embodiment, the first clearing model for power demand side flexibility sharing is constructed based on the objective function, coupled constraint condition and non-coupled constraint condition; the coupled constraint condition includes supply-demand balance constraint condition; the non-coupled constraint condition includes energy regulation constraint condition, supply vector direction constraint condition and user individual feasible region constraint condition.

[0055] The first clearing model is:

[0056] ;

[0057] Wherein, is the flexibility supply vector of the user , and are the regulation cost coefficients of the user in the time period , is the further regulation power of the user compared with the self-regulation power curve regulation cost generated to provide flexibility, total demand for flexibility sharing, Lagrange multiplier for supply-demand balance constraint condition, for users individual feasible region of users, for Hadamard product operation.

[0058] Specifically, the first clearing model is constructed by an objective function , a supply-demand balance constraint condition in a coupling constraint condition , an energy regulation constraint condition in a non-coupling constraint condition , a supply vector direction constraint condition , a user individual feasible region constraint condition , and is reported by users autonomously.

[0059] Step S12: based on the power demand side flexibility sharing first clearing model, a power demand side flexibility sharing second clearing model is constructed, the power demand side flexibility sharing second clearing model is solved to output a current iteration initial value; the current iteration initial value includes a Lagrange multiplier and a flexibility supply average value vector.

[0060] In the embodiment, the power demand side flexibility sharing second clearing model is constructed based on the objective function, the supply-demand balance constraint condition, the energy regulation constraint condition and the supply vector direction constraint condition.

[0061] The way that the first clearing model is cleared by the organizer in a centralized manner is not conducive to the protection of the privacy of participating members on the one hand, and the expansion of the scale of flexibility sharing participants will lead to heavy computing burden, which is not conducive to expansion to accommodate a large number of DR (Demand response, power demand response) users. The alternating direction multiplier method decomposes the optimization problem with separable variables into multiple relatively independent sub-problems, solves the dual variables through iteration update, has high convergence and robustness, and does not require the objective function to be strictly convex, and has important application in distributed optimization and dispatching of power systems. The above optimization problem is solved by using a distributed clearing method based on ADMM. Although the iteration initial value is directly taken as 0, it can converge to the optimal solution, but the iteration number is large and the convergence time is slow. In order to speed up the solving process, the hot start strategy is designed as follows: the present application proposes a hot start method that takes the optimal solution of the clearing model without considering user individual constraints as the iteration initial value, which greatly speeds up the convergence speed of the distributed clearing algorithm while protecting user privacy, and realizes the accelerated solution of distributed clearing. A second clearing model without considering user individual constraints is defined as follows:

[0062] ;

[0063] Note the second clearing model The optimal solution is The Lagrange multiplier for the supply-demand balance constraint of flexibility sharing under the optimal value condition is: Compared to the first clearing model This optimization problem Disregarding individual user constraints Therefore, without requiring users to report individual operational constraints, the organizer can directly model and solve the problem based on the bidding information. It should be noted that the solution to this optimization problem... This is only the first clearing model. The "suboptimal solution". This occurs when the user has no individual constraints or the individual constraints are ineffective. This is the first clearing model. The optimal solution. Therefore, the initial values ​​for the distributed clearing iteration can be given, where, This is the initial value for the ADMM algorithm iteration. As a penalty factor:

[0064] .

[0065] In this application, flexibility refers to demand-side flexibility, which means that users participating in DR can adjust their own electricity load to track the ideal load curve target published by the grid and support grid regulation. Flexibility sharing is the way that electricity users with different flexibility levels can track the ideal target curve published by the demand response through trading flexibility.

[0066] In this embodiment, the process of calculating the Lagrange multipliers is as follows: constructing an augmented Lagrange function based on the objective function and coupling constraints; and calculating the Lagrange multipliers using the augmented Lagrange function.

[0067] The augmented Lagrange function is:

[0068] ;

[0069] in, For users The flexibility of the supply vector, For the Lagrange multipliers corresponding to the supply and demand balance constraints, , To allow users to further adjust power compared to the self-adjusted power curve. The adjustment costs arising from providing flexibility, To meet the overall need for flexible sharing, As a penalty factor, .

[0070] The energy regulation constraints, the supply vector direction constraints, and the individual feasible region constraints of users are only related to the flexible supply of the user , and are independent of the flexible supply of other users, which are called uncoupled constraints. The flexible supply-demand balance constraints are called coupled constraints. The feasible region composed of uncoupled constraints is denoted as . The augmented Lagrangian function of the clear model can be obtained according to the objective function and the coupled constraints.

[0071] The standard alternating iteration form based on the ADMM algorithm is as follows:

[0072] ;

[0073] where k is the iteration round. The update of the flexible supply variable is completed independently by the user , and only needs to consider the uncoupled constraints, that is, optimization is performed in the feasible region . In addition, when the flexible supply variable is updated in the kth iteration, the value of the flexible supply variable updated in the kth iteration is taken , the value of the flexible supply variable updated in the k-1th iteration is taken. After the update of is completed, the Lagrange multiplier is updated. The iteration is stopped until the convergence condition is met.

[0074] Let , the above formula can be simplified as:

[0075] ;

[0076] where the update of the variable depends on the latest updated values of the remaining variables, and requires the exchange of the updated optimization variables between users. On the one hand, users usually lack channels for information transmission, and it is difficult to update the variables by interacting with each other. On the other hand, the exchange of variable update values is also not conducive to the protection of user privacy. In order to overcome the above problems caused by information exchange between users, a synchronous iteration method is proposed to update the optimization variables of users in each iteration.

[0077] In this embodiment, the formula for calculating the average flexible supply vector is as follows:

[0078] ;

[0079] where represents the average flexible supply vector after the kth iteration. ​

[0080] Step S13: Calculate the flexibility demand surplus by using the current iteration initial value, and send the current iteration initial value and the flexibility demand surplus to the client, so that the client calculates the flexibility supply variable based on the current iteration initial value and the flexibility demand surplus on the basis of the user individual feasible region constraint condition.

[0081] In this embodiment, the process of calculating the flexibility demand surplus is as follows: the current iteration initial value is calculated by using the flexibility demand surplus calculation formula, and the flexibility demand surplus is obtained.

[0082] The flexibility demand surplus calculation formula is as follows:

[0083] ;

[0084] Among them, is the flexibility demand surplus after the kth iteration, which represents the imbalance of the supply and demand of flexibility sharing, is the flexibility supply variable. It is not difficult to see that when is 0, there is no imbalance between supply and demand, and the flexibility supply exactly meets the total demand.

[0085] Step S14: Update the flexibility supply average value vector and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client, and obtain the updated flexibility demand surplus.

[0086] Step S15: Update the Lagrange multiplier by using the updated flexibility demand surplus, and obtain the updated Lagrange multiplier.

[0087] Replace in the above simplified formula with and , and the following ADMM-based synchronous iteration method can be obtained in combination with the flexibility demand surplus calculation formula:

[0088] ;

[0089] ;

[0090] In the kth iteration, the organizer calculates and based on the iteration results of the k-1th iteration, and publishes them to the users . The users update the flexibility supply variable based on the above formula, and finally the organizer updates the Lagrange multiplier .

[0091] Step S16: judging whether the updated Lagrange multiplier and the average value vector of the flexibility supply meet a preset convergence condition, if the updated Lagrange multiplier and the average value vector of the flexibility supply meet the preset convergence condition, generating a clearing result based on the flexibility supply variable; if the updated Lagrange multiplier and the average value vector of the flexibility supply do not meet the preset convergence condition, jumping to the process of calculating the flexibility demand surplus by using the current iteration initial value, returning the current Lagrange multiplier, the flexibility demand surplus and the average value of the flexibility supply to the client, so as to update the flexibility supply variable by the client.

[0092] In the embodiment, it is judged whether the norm square of the original residual of the updated Lagrange multiplier is not greater than a preset original residual convergence threshold value; and it is judged whether the norm square of the dual residual of the average value vector of the flexibility supply is not greater than a preset dual residual convergence threshold value; if the updated Lagrange multiplier and the average value vector of the flexibility supply meet the preset convergence condition, generating a clearing result based on the flexibility supply variable; if the updated Lagrange multiplier and the average value vector of the flexibility supply do not meet the preset convergence condition, updating the current iteration number, returning the current Lagrange multiplier, the flexibility demand surplus and the average value of the flexibility supply to the client, so as to calculate the updated flexibility supply variable by the client based on the current Lagrange multiplier, the flexibility demand surplus and the average value of the flexibility supply returned to the client; jumping to the process of updating the average value vector of the flexibility supply and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client, until the updated Lagrange multiplier and the average value vector of the flexibility supply meet the preset convergence condition.

[0093] Specifically, the preset convergence condition is that the l-2 norm square of the original residual and the dual residual is less than the set convergence threshold value:

[0094] ;

[0095] wherein, is the convergence threshold value of the original residual, is the convergence threshold value of the dual residual.

[0096] Step S17: realizing the sharing transaction of the power demand side distributed clearing by using the clearing result.

[0097] The distributed clearing process of the flexibility sharing mechanism based on ADMM is as shown in Figure 2 , and the specific steps are as follows:

[0098] (1) Let the iteration number k=0; initialize and , and the process uses the formula ;

[0099] (2) Calculation , the process uses the formula ; k = 1;

[0100] (3) User decision and report , the process uses the formula ;

[0101] (4) The organization side updates and , and publishes to the user, the process uses the formula , ;

[0102] (5) The organization side updates , and publishes to the user, ;

[0103] (6) The organization side judges whether the convergence condition is met : if met, the iteration ends and the clear result is output; if not met, k = k + 1, return to step (3) until the convergence condition is met.

[0104] In view of the fact that centralized clearing is not conducive to user privacy protection and the organization side has heavy computing burden, the distributed solution based on ADMM has many iterations and slow convergence, and the existing accelerated clearing method is difficult to apply to flexible sharing distributed clearing, the present application proposes a warm start accelerated solution method for power demand side flexible sharing distributed clearing. On the basis of introducing a synchronous ADMM algorithm, a second clearing model is constructed, and the optimization result of the second clearing model is used as the initial value of iteration to reduce the redundant steps of the original ADMM iterative solution from zero, significantly reduce the iteration number and convergence time under the premise of protecting user privacy, improve the speed of distributed solution, and realize efficient distributed solution of the flexible sharing clearing model. The method not only does not need the user to report individual operation constraints, avoids the leakage of user privacy, but also distributes the computing pressure to individual users, can adapt to large-scale user participation in flexible sharing, support large-scale development of flexible sharing, and promote the flexible use of demand side.

[0105] The effectiveness of the flexible sharing method is illustrated by a specific embodiment: a system with 4 DR users is used for analysis. The purchase curve of the 4 users under the condition of no DR according to their own power consumption habits or economic benefit maximization is shown in Figure 3 The quadratic cost coefficient of the flexible supply of the four users is As shown in Table 1, the linear cost coefficient of the flexible supply is As shown in Table 2:

[0106] Table 1 Quadratic cost coefficients of flexible supply

[0107]

[0108] Table 2 Linear cost coefficients of flexible supply

[0109]

[0110] The constraint of individual load adjustment is that user 1 cannot exceed the original electricity consumption, user 2 and user 3 are limited within ±10% of the average load, and user 4 is limited within ±50% of the average load. The highest price of the flexible sharing mechanism is set to 15 yuan / kWh. The penalty factor The original residual convergence threshold is 0.03 The dual residual convergence threshold is . .

[0111] Based on the scheme, the convergence process of distributed clearing can be calculated, Figure 4 is the iterative convergence process diagram of the actual value and the logarithmic value with base 10 and the original residual, wherein, Figure 4 (a) in FIG. 4 is the iterative convergence process diagram of the actual value and the original residual, Figure 4 (b) in FIG. 4 is the iterative convergence process diagram of the logarithmic value with base 10 and the original residual; Figure 5 is the iterative convergence process diagram of the actual value and the logarithmic value with base 10 and the dual residual, wherein, Figure 5 (a) in FIG. 5 is the iterative convergence process diagram of the actual value and the dual residual, Figure 5 (b) in FIG. 5 is the iterative convergence process diagram of the logarithmic value with base 10 and the dual residual. It can be seen that 166 iterations are required to meet the convergence condition without hot start, and only 36 iterations are required with hot start, which greatly improves the solving efficiency. The reason is that the centralized clearing result without considering individual constraints is closer to the global optimal solution than the 0 initial point or the random initial point, which provides a better starting point for ADMM iteration, thereby reducing the distance of searching for a solution in the search space and reducing the number of iterations. The comparison of the flexible sharing cost of centralized and distributed clearing is shown in FIG. 6 Figure 6 Figure 6 The comparison of the flexible supply cost of centralized and distributed clearing is shown in FIG. 7. The cost tends to be stable after 150 iterations without hot start, and the cost tends to be stable after more than 30 iterations with hot start, which is consistent with the Figure 4 , Figure 5 result. Compared with the result of centralized clearing, the error of flexible supply cost is 0.07% and 0.023% respectively. It shows that distributed clearing can effectively approximate the result of centralized clearing within a small error range, and hot start can further reduce the error.​​

[0112] Therefore, the scheme based on the hot start acceleration distributed clearing can realize efficient and accurate solution of the flexibility sharing model, efficiently obtains the flexibility sharing transaction result under the premise of protecting the user, and is beneficial to large-scale user participation in demand side flexibility sharing.

[0113] The application has the advantages that: by using the optimization result of the approximate clearing model as the initial iteration value, the redundant steps of the original ADMM from zero start iteration are reduced, the iteration number of the ADMM algorithm is effectively reduced, the convergence time is shortened, and the solving efficiency is significantly improved; without the user reporting individual operation constraints, the privacy protection problem caused by centralized information processing in centralized solving is eliminated, and the risk of user privacy leakage is effectively avoided; the calculation burden of the organization party can be reduced, since the calculation pressure is distributed to the individual users, the over-reliance of centralized solving on the calculation resources of the organization party is avoided, the large-scale user participation scenario can be adapted, and the large-scale development of flexibility sharing can be supported; a variety of distributed resources and non-convex problems can be adapted, the individual decision problem given in the technical content does not require strict convexity, the user does not necessarily make decisions according to the optimization model, the decision result can be given according to the actual situation considering the self-operation cost, bounded rationality and other factors, and the scenario of a plurality of types of distributed resources (such as wind turbines, photovoltaic, electric vehicles and the like) participating in sharing which is difficult to establish accurate operation model constraints can be well adapted.

[0114] The application considers the solving efficiency, calculation burden and adaptability on the basis of protecting the user privacy, is an important improvement and innovation over the prior art, can adapt to large-scale and complex distributed flexibility sharing transaction scenarios, and supports effective mining and utilization of demand side flexibility.

[0115] In the embodiment, a first clearing model of power demand side flexibility sharing is constructed; a second clearing model of power demand side flexibility sharing is constructed based on the first clearing model of power demand side flexibility sharing, and the second clearing model of power demand side flexibility sharing is solved to output a current iteration initial value; the current iteration initial value includes a Lagrange multiplier and a flexibility supply average value vector; a flexibility demand surplus is calculated by using the current iteration initial value, and the current iteration initial value and the flexibility demand surplus are sent to a client, so that the client calculates a flexibility supply variable based on the current iteration initial value and the flexibility demand surplus on the basis of a user individual feasible domain constraint condition; the flexibility supply average value vector and the flexibility demand surplus are updated based on the flexibility supply variable updated and sent by the client to obtain an updated flexibility demand surplus; the Lagrange multiplier is updated by using the updated flexibility demand surplus to obtain an updated Lagrange multiplier; it is judged whether the updated Lagrange multiplier and the flexibility supply average value vector meet a preset convergence condition, if the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition, a clearing result is generated based on the flexibility supply variable; if the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, the process of calculating the flexibility demand surplus by using the current iteration initial value is jumped to, and the current Lagrange multiplier, the flexibility demand surplus and the flexibility supply average value are returned to the client, so that the client updates the flexibility supply variable; and the sharing transaction of power demand side distributed clearing is realized by using the clearing result.The application aims at the problems of centralized clearing being not conducive to user privacy protection, the heavy computing burden of organizers, the large number of iterations and slow convergence of distributed solution based on alternating direction multiplier algorithm, and the difficulty of existing accelerated clearing methods to be applied to flexible sharing distributed clearing. A power demand side flexible sharing first clearing model is constructed, a power demand side flexible sharing second clearing model is constructed based on the power demand side flexible sharing first clearing model, the power demand side flexible sharing second clearing model is solved to output a current iteration initial value, the current iteration initial value includes a Lagrange multiplier and a flexible supply average value vector, which can reduce the redundant steps of iteration solution from zero of the alternating direction multiplier algorithm, the flexible demand surplus is calculated by using the current iteration initial value, and the current iteration initial value and the flexible demand surplus are sent to the client, so that the client calculates the flexible supply variable based on the current iteration initial value and the flexible demand surplus on the basis of the user individual feasible region constraint condition, updates the flexible supply average value vector and the flexible demand surplus based on the flexible supply variable, obtains the updated flexible demand surplus, updates the Lagrange multiplier by using the updated flexible demand surplus, obtains the updated Lagrange multiplier, judges whether the updated Lagrange multiplier and the flexible supply average value vector meet a preset convergence condition, if the preset convergence condition is met, a clearing result is generated based on the flexible supply variable, distributed clearing solution is realized, large-scale user participation in flexible sharing is adapted, and the accuracy and efficiency of shared transactions are improved.

[0116] Figure 7 A structural schematic diagram of an electronic device is provided for the embodiments of the application. The electronic device 20 can specifically include at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is configured to store a computer program, the computer program is loaded and executed by the processor 21 to implement the related steps in the power demand side flexible sharing transaction hot start distributed accelerated clearing method executed by the electronic device disclosed in any of the preceding embodiments.

[0117] In the embodiment, the power supply 23 is configured to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol followed by the communication interface 24 can be any communication protocol applicable to the technical solution of the application, which is not limited specifically herein; the input / output interface 25 is configured to obtain external input data or output data to the outside, and the specific interface type can be selected according to specific application needs, which is not limited specifically herein.

[0118] In addition, the memory 22 can be a read-only memory, a random access memory, a magnetic disk or an optical disk, etc. as a carrier for storing resources, and the resources stored thereon include an operating system 221, a computer program 222, data 223, etc. The storage mode can be temporary storage or permanent storage.

[0119] The operating system 221 is used to manage and control each hardware device on the electronic device 20 and the computer program 222, so as to realize the operation and processing of the processor 21 on the data 223 in the memory 22. The operating system 221 can be Windows, Unix, Linux, etc. In addition to the computer program capable of completing the hot start distributed acceleration clearing method of the power demand side flexibility sharing transaction executed by the electronic device 20 disclosed in any one of the preceding embodiments, the computer program 222 can further include a computer program capable of completing other specific work. In addition to the data transmitted by the external device and received by the power demand side flexibility sharing transaction hot start distributed acceleration clearing device, the data 223 can also include the data collected by the input and output interface 25 itself and the like.

[0120] The steps of the method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module can be located in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0121] Further, the embodiments of the present application also disclose a computer readable storage medium, wherein the storage medium stores a computer program, and the computer program is loaded and executed by a processor to realize the steps of the hot start distributed acceleration clearing method of the power demand side flexibility sharing transaction disclosed in any one of the preceding embodiments.

[0122] Finally, it should be noted that, in this document, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, article or device including the element.

[0123] The above describes in detail the power demand side flexible sharing transaction hot start distributed acceleration out clearing method, equipment and storage medium provided by the present application. The principles and implementation modes of the present application are described by using specific examples. The above example is only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed. In summary, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A hot start distributed accelerated discharge method for power demand side flexibility sharing transaction, characterized in that, The method comprises the following steps: constructing a first clearing model of power demand side flexibility sharing; constructing a second clearing model of power demand side flexibility sharing based on the first clearing model of power demand side flexibility sharing, and solving the second clearing model of power demand side flexibility sharing to output a current iteration initial value; the current iteration initial value comprises a Lagrange multiplier and a flexibility supply average value vector; calculating a flexibility demand surplus by using the current iteration initial value, and sending the current iteration initial value and the flexibility demand surplus to a client, so that the client calculates a flexibility supply variable based on the current iteration initial value and the flexibility demand surplus on the basis of a user individual feasible region constraint condition; updating the flexibility supply average value vector and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client to obtain an updated flexibility demand surplus; updating the Lagrange multiplier by using the updated flexibility demand surplus to obtain an updated Lagrange multiplier; judging whether the updated Lagrange multiplier and the flexibility supply average value vector meet a preset convergence condition; if the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition, generating a clearing result based on the flexibility supply variable; if the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, jumping to the process of calculating the flexibility demand surplus by using the current iteration initial value, and returning the current Lagrange multiplier, the flexibility demand surplus and the flexibility supply average value to the client, so that the client updates the flexibility supply variable; realizing sharing transaction of power demand side distributed clearing by using the clearing result; the first clearing model is as follows: ； in, For users The flexibility of the supply vector, and User exist Adjustment cost coefficient for different time periods To allow users to further adjust power compared to the self-adjusted power curve. The adjustment costs arising from providing flexibility, To meet the overall need for flexible sharing, For the Lagrange multipliers under supply and demand equilibrium constraints, For users Individual operating domain, This is the Hadama product operation. 2.The thermal start-up distributed acceleration discharge method of power demand side flexibility sharing transaction according to claim 1, wherein, the process of constructing the first clearing model of power demand side flexibility sharing comprises the following steps: constructing the first clearing model of power demand side flexibility sharing based on an objective function, coupled constraint conditions and non-coupled constraint conditions; the coupled constraint conditions comprise a supply-demand balance constraint condition; the non-coupled constraint conditions comprise an energy regulation constraint condition, a supply vector direction constraint condition and a user individual feasible region constraint condition; correspondingly, the process of constructing the second clearing model of power demand side flexibility sharing comprises the following steps: constructing the second clearing model of power demand side flexibility sharing based on the objective function, the supply-demand balance constraint condition, the energy regulation constraint condition and the supply vector direction constraint condition. 3.The hot start distributed acceleration out-clearing method of power demand side flexibility sharing transaction according to claim 1, wherein, the process of calculating the Lagrange multiplier comprises the following steps: constructing an augmented Lagrange function based on an objective function and coupled constraint conditions; calculating the Lagrange multiplier by using the augmented Lagrange function. 4.The hot start distributed acceleration discharge method of power demand side flexibility sharing transaction according to claim 3, wherein, the augmented Lagrange function is as follows: ； wherein, is the flexibility supply vector for the users, is the flexibility demand vector for the users, is the Lagrange multiplier corresponding to the supply-demand balance constraint, , is the further adjustment of the power for the users compared to the self-adjusted power curve is the adjustment cost resulting from the provision of flexibility, is the total demand for flexibility sharing, is the penalty factor, . 5.The thermal start-up distributed acceleration discharge method of power demand side flexibility sharing transaction according to claim 1, wherein, the process of calculating the flexibility demand surplus by using the current iteration initial value comprises the following steps: calculating the flexibility demand surplus by using a flexibility demand surplus calculation formula on the current iteration initial value to obtain the flexibility demand surplus; the flexibility demand surplus calculation formula is as follows: ； wherein, is the remaining flexibility demand after the kth iteration, is the flexibility supply variable. 6.The thermal start-up distributed acceleration discharge method of power demand side flexibility sharing transaction according to claim 1, wherein, the process of judging whether the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition comprises the following steps: judging whether a norm square of an original residual error of the updated Lagrange multiplier is not greater than a preset original residual error convergence threshold. And, judge whether the norm square of the dual residual of the flexibility supply average value vector is not greater than a preset dual residual convergence threshold. 7.The hot start distributed acceleration discharge method of power demand side flexibility sharing transaction according to any one of claims 1 to 6, characterized in that, If the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, jump to a process of calculating the flexibility demand surplus by using the current iteration initial value, return the current Lagrange multiplier, the flexibility demand surplus, and the flexibility supply average value to the client, so that the client updates the flexibility supply variable, including: If the updated Lagrange multiplier and the flexibility supply average value vector do not meet the preset convergence condition, update the current iteration number, return the current Lagrange multiplier, the flexibility demand surplus, and the flexibility supply average value to the client, so that the client calculates the updated flexibility supply variable based on the current Lagrange multiplier, the flexibility demand surplus, and the flexibility supply average value returned to the client; Jump to the process of updating the flexibility supply average value vector and the flexibility demand surplus based on the flexibility supply variable updated and sent by the client until the updated Lagrange multiplier and the flexibility supply average value vector meet the preset convergence condition.

8. An electronic device, comprising: Comprise: A memory for saving a computer program; A processor for executing the computer program to implement the hot start distributed acceleration out-of-clear method for power demand side flexibility sharing transaction according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, A memory for saving a computer program; wherein the computer program is executed by a processor to implement the hot start distributed acceleration out-of-clear method for power demand side flexibility sharing transaction according to any one of claims 1 to 7.

Images