Full privacy protection energy scheduling method and device
Through random sparse non-singular matrix, the parameter matrix of the electric-thermal coupling system is transformed in the primary transformation and the internal point method is used to solve the problem of large overhead of private information leakage and encryption calculation in the power system and regional heating systems, and the information is realized The balance between safety and result accuracy improves the scheduling synergy and energy consumption efficiency of the electrical-thermal coupled multi-energy flow system.
Patent Information
- Application Number
- CN202411406918.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-10
- Publication Date
- 2025-01-07
AI Technical Summary
There is a risk of private information leakage in the existing power system and regional heating system in the distributed scheduling of electricity-thermal coupled, and the existing encryption methods have high computational overhead and difficult to ensure the accuracy of the results, especially in large-scale scenario applications. .
Random sparse non-singular matrix is used to perform elementary transformation of the parameter matrix in the scheduling model system, and iteratively solve the encrypted model system with the inner point method to ensure the security of private information and the accuracy of the results.
While not increasing the difficulty of calculation, it realizes effective encryption of private information and the accuracy of results, improves the scheduling synergy and energy consumption efficiency of the electrical-thermal coupled multi-energy flow system, and reduces operating costs.
Smart Images
Figure CN119273086A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of power grid operation and control technology in the form of energy, and in particular to a full privacy protection energy scheduling method and device. Background Art
[0002] At present, the power system and the district heating system are operated based on the distributed scheduling method of the power-heat coupling system in engineering practice. Although the distributed algorithm only requires limited data exchange when promoting the collaboration between the power system and the district heating system, there is still a risk of private information leakage. Among them, private information includes parameter information related to the collaborative scheduling model of the power-heat coupling system, such as economic parameters in the objective function and physical characteristics in the constraints. For example: on the one hand, by collecting the information exchanged in each iteration, and further using big data, deep learning and other technologies to infer private information. On the other hand, the exchanged information may be stolen from the communication channel, such as the output of the cogeneration unit and the gradient of the boundary constraint.
[0003] Existing encryption methods are often based on homomorphic encryption or differential privacy methods. Although they can ensure the security of private information, they often have defects such as difficulty in ensuring the accuracy of results, high computational overhead of encryption problems, and difficulty in large-scale application scenarios. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a fully privacy-protected energy scheduling method and device, which can ensure the security of private information and the accuracy of results without greatly increasing the difficulty of calculation.
[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0006] A fully privacy-preserving energy scheduling method, comprising:
[0007] Obtain the scheduling model system to be encrypted;
[0008] Performing elementary transformation on the parameter matrix in the scheduling model system by using a random sparse non-singular matrix to obtain an encrypted scheduling model system;
[0009] Initialize the number of iterations and solve the encryption scheduling model system by an interior point method;
[0010] Determine whether the optimal objective function value corresponding to the current encrypted scheduling model system is a preset value. If so, converge to the global optimal solution to obtain the scheduling plan corresponding to the scheduling model system.
[0011] In order to solve the above technical problems, another technical solution adopted by the present invention is:
[0012] A fully privacy-protected energy scheduling device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step in the above-mentioned fully privacy-protected energy scheduling method is implemented.
[0013] The beneficial effects of the present invention are as follows: by performing elementary transformation on the parameter matrix of the scheduling problem in the scheduling model system through a random sparse non-singular matrix, the information in the parameter matrix is encrypted, which can ensure the security of private information and the accuracy of the results without greatly increasing the difficulty of calculation; and then the encrypted scheduling model system is iteratively solved through the interior point method to obtain the optimal scheduling solution. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 is a flowchart of the steps of a fully privacy-protected energy scheduling method in an embodiment of the present invention;
[0015] Figure 2 Schematic diagram of the structure of a fully privacy-protected energy scheduling device in an embodiment of the present invention. DETAILED DESCRIPTION
[0016] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.
[0017] Please refer to Figure 1 , a fully privacy-preserving energy scheduling method, including:
[0018] Obtain the scheduling model system to be encrypted;
[0019] Performing elementary transformation on the parameter matrix in the scheduling model system by using a random sparse non-singular matrix to obtain an encrypted scheduling model system;
[0020] Initialize the number of iterations and solve the encryption scheduling model system by an interior point method;
[0021] Determine whether the optimal objective function value corresponding to the current encrypted scheduling model system is a preset value. If so, converge to the global optimal solution to obtain the scheduling plan corresponding to the scheduling model system.
[0022] From the above description, it can be seen that the beneficial effects of the present invention are: by performing elementary transformation on the parameter matrix of the scheduling problem in the scheduling model system through a random sparse non-singular matrix, the information in the parameter matrix is encrypted, which can ensure the security of private information and the accuracy of the results without greatly increasing the difficulty of calculation; and then the encrypted scheduling model system is iteratively solved by the interior point method to obtain the optimal scheduling solution.
[0023] Furthermore, the scheduling model system includes a heating system scheduling model and a power system scheduling model;
[0024] The initializing the number of iterations and solving the encryption scheduling model system by the interior point method includes:
[0025] The encrypted power system dispatch model is solved by using the interior point method to obtain power supply dispatch parameters;
[0026] Based on the power supply scheduling parameters, the encrypted heating system scheduling model is solved using the interior point method to obtain the optimal objective function value corresponding to the heating system scheduling model.
[0027] From the above description, it can be seen that the method in this embodiment can be applied to the scheduling plan formulation of the electric-thermal coupled multi-energy flow system. After first solving the power system scheduling model to obtain the power supply scheduling parameters, the encrypted heating system scheduling model is solved based on the power supply scheduling parameters, thereby realizing the coordinated scheduling of the electric-thermal coupled multi-energy flow system.
[0028] Furthermore, the power system dispatch model includes:
[0029]
[0030] Among them, p is the dispatch decision variable in the power system dispatch model, ξ is a non-negative auxiliary variable, is a non-negative slack variable; a and b are the coefficient matrices corresponding to the objective function; E p and F p is the coefficient matrix corresponding to the constraints, U p is the coefficient matrix associated with the slack variables;
[0031] The power supply system dispatcher generates a random sparse non-singular matrix Y p and I p (To avoid a significant increase in computational difficulty, Y p and I p The density of the matrix is not greater than 0.5%). The elementary transformation of the parameter matrix in the scheduling model system includes:
[0032]
[0033] Among them, Y p and I p represents a sparse non-singular matrix;
[0034] From the above description, it can be seen that the established power system dispatching model is expressed in matrix form without loss of generality, so that the subsequent process can perform corresponding encryption changes based on the matrix form to achieve encryption of private information in the parameter matrix.
[0035] Furthermore, the heating system scheduling model includes:
[0036]
[0037] Among them, α, h and τ are variable vectors representing the parameters in the heating system scheduling model, s h are the slack variables, c, E, U, G (ψ∈{p,α,h,τ}) are the corresponding parameter matrices in the objective function and constraints in the heating system scheduling model;
[0038] The elementary transformation of the parameter matrix in the scheduling model system by using a random sparse non-singular matrix includes:
[0039]
[0040] in, It is the boundary power information transmitted from the power supply system to the heating system.
[0041]
[0042] From the above description, it can be seen that the established heating system scheduling model is expressed in matrix form without loss of generality, so that the subsequent process can perform corresponding encryption changes based on the matrix form to achieve encryption of private information in the parameter matrix.
[0043] Furthermore, before solving the encrypted heating system scheduling model based on the power supply scheduling parameters by using the interior point method, the method further includes:
[0044] The encrypted heating system scheduling model is equivalently expressed as a feasibility problem to obtain a feasibility model;
[0045] A normalization condition is obtained, and the normalization condition is combined with the dual problem of the feasibility model to obtain a minimum infeasibility problem model.
[0046] From the above description, it can be seen that the encrypted heating system scheduling model is equivalently expressed as a feasibility problem and solved in combination with normalization conditions. Compared with the traditional encryption method, the distributed solution performance of the algorithm can be greatly improved.
[0047] Furthermore, the feasibility model includes:
[0048]
[0049] in, Boundary cost information transferred from the power supply system to the heating system.
[0050] From the above description, it can be seen that by introducing the non-negative auxiliary variable ξ, the encrypted heating system scheduling model is equivalently expressed as a feasibility problem, which can improve the distributed solution performance of the algorithm.
[0051] Furthermore, the normalization condition includes:
[0052]
[0053] The minimum infeasibility problem model includes:
[0054]
[0055] in,
[0056]
[0057] and Respectively The dual multipliers corresponding to each constraint in , and Respectively and The corresponding dual multiplier.
[0058] From the above description, it can be seen that by combining the normalization condition with the dual problem of the feasibility model to obtain the minimum infeasible problem model, the distributed solution performance of the algorithm can be improved.
[0059] Furthermore, the determining whether the optimal objective function value corresponding to the current encryption scheduling model system is a preset value also includes:
[0060] If so, it converges to the global optimal solution to obtain the scheduling solution corresponding to the scheduling model system; for example, if the optimal objective function value of the minimum infeasibility model is equal to 0, it converges to the global optimal solution use Decryption can obtain the optimal collaborative scheduling solution If not, the maximum non-dominated cutting plane inequality is generated according to the encrypted heating system scheduling model and power supply scheduling parameters, specifically:
[0061]
[0062] The encrypted scheduling model system is re-solved based on the maximum non-dominated cut plane inequality as a constraint condition.
[0063] From the above description, it can be seen that by solving the minimum infeasibility problem to generate the maximum non-dominated cutting plane, that is, the Pareto optimal cut rather than the ordinary cutting plane, the distributed solving performance of the algorithm can be further improved.
[0064] Another embodiment of the present invention provides a fully privacy-protected energy scheduling device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements each step in a fully privacy-protected energy scheduling method as described above when executing the computer program.
[0065] The energy scheduling method and device with full privacy protection provided by the present invention can be applied to the scenario of scheduling plan formulation of electric-thermal coupled multi-energy flow system, and can be adapted to the original power system and regional heating system energy management system, which is described below through specific implementation methods:
[0066] Embodiment 1
[0067] Please refer to Figure 1 , a fully privacy-preserving energy scheduling method, including:
[0068] S1. Obtain a scheduling model system to be encrypted. In this embodiment, a distributed scheduling method for an electric-thermal coupling system is used for explanation. The scheduling model system includes a heating system scheduling model and a power system scheduling model.
[0069] S11. The construction of the power system dispatch model includes: the objective function and constraints of the power system dispatch model. The constraints include the DC power flow equation constraints in the power system, the active power constraints of non-CHP generator sets, the active power constraints of wind turbine sets, the ramping constraints of the active power of non-CHP generator sets, and the power system backup constraints, etc. It also includes that according to the actual industrial operation scenario, the system operation may also need to meet other linear constraints, such as energy storage charging and discharging constraints, carbon emission constraints, etc.;
[0070] The objective function is expressed as:
[0071]
[0072] Among them, κ WD represents the wind turbine set, κ TU represents the set of non-CHP generators, T is the set of scheduling periods; is the power generation cost of the g-th non-CHP generator unit in period t, is the power generation cost of the w-th wind turbine in period t, are the cost constant term coefficient, first-order term coefficient and second-order term coefficient of the g-th non-CHP generator set, which can be obtained from the factory manual of the non-CHP generator set. w is the cost coefficient of the w-th wind turbine, which can be obtained from the price specified in the electricity market. represents the maximum available power output of the wth wind turbine in period t, which can be obtained from the wind power prediction module; p g,trepresents the power output of the g-th unit in period t, It represents the actual power output of the wth wind turbine in period t.
[0073] After constructing the power system dispatch model, the established power system dispatch model is expressed in matrix form without loss of generality:
[0074]
[0075] Among them, p is the dispatch decision variable in the power system dispatch model, ξ is a non-negative auxiliary variable, A non-negative slack variable introduced to convert an inequality constraint into an equality constraint (a slack variable is essentially a margin that indicates when the inequality is not "tight"; when the inequality is strictly less than Non-zero; when the inequality reaches the boundary (that is, when the equality sign holds), the relaxation variable );a and b are the coefficient matrices corresponding to the objective function; E p and F p is the coefficient matrix corresponding to the constraints, U p is the coefficient matrix associated with the slack variables, which is a diagonal matrix consisting of 0 and 1 elements. A value of 1 indicates that the corresponding inequality constraint is relaxed, and a value of 0 indicates that the original constraint is an equality constraint.
[0076] S12. Construction of the heating system scheduling model includes: objective function and constraints of the heating system scheduling model; wherein the constraints include operating characteristic equation constraints of heating equipment (including CHP units and boilers) in the heating system, power output constraints of thermal power plants, heat exchange equation constraints of heat sources, water supply temperature constraints of heat sources, temperature equation constraints of the junction of multiple pipes in the heating network, temperature correlation equation constraints of the heating network, heat loss equation constraints of the heating network pipelines, heat exchange equation constraints of loads, and return water temperature constraints of heat loads. In addition, according to the actual industrial operation scenario, the system operation may also need to meet other linear constraints, such as heat storage constraints, electric boiler constraints, etc.
[0077] The objective function is expressed as:
[0078]
[0079] Among them, κ CHP represents the set of CHP units, κ HB represents the boiler set, c g,t is the cost of the g-th device in period t.
[0080] After constructing the heating system scheduling model, the established heating system scheduling model is expressed in matrix form without loss of generality:
[0081]
[0082] Among them, α, h and τ are variable vectors representing the parameters in the heating system scheduling model, s h are the slack variables, c, E, U, and G (ψ∈{p,α,h,τ}) are the corresponding parameter matrices in the objective function and constraints of the heating system scheduling model, respectively.
[0083] S2, performing elementary transformation on the parameter matrix in the scheduling model system through a random sparse non-singular matrix to obtain an encrypted scheduling model system;
[0084] S21. For the power system dispatching model, the power supply system dispatcher generates a random sparse non-singular matrix Y p and I p (To avoid a significant increase in computational difficulty, Y p and I p The density of the matrix is not greater than 0.5%), and the model constructed in step S11 is encrypted to obtain the power system encrypted scheduling model:
[0085]
[0086] Among them, Y p and I p represents a sparse non-singular matrix;
[0087] S22. For the heating system scheduling model, the heating system scheduling center generates a random positive definite sparse matrix Y ψ and I ψ (ψ∈{α,h,τ})(To avoid a significant increase in computational difficulty, Y ψ and I ψ The density of the matrix is not greater than 0.5%), and the model constructed in step S12 is encrypted to obtain an encrypted scheduling model of the heating system:
[0088]
[0089] in, It is the boundary power information transmitted from the power supply system to the heating system.
[0090]
[0091] S23, the encrypted heating system scheduling model (i.e., the model in step S22) is equivalently expressed as a feasibility problem to obtain a feasibility model; wherein the feasibility model includes:
[0092]
[0093] in, Boundary cost information transferred from the power supply system to the heating system.
[0094] S24, obtaining a normalization condition, and combining the normalization condition with the dual problem of the feasibility model to obtain a minimum infeasibility problem model; wherein the normalization condition includes:
[0095] The minimum infeasibility problem model includes:
[0096]
[0097] in,
[0098]
[0099] and Respectively The dual multipliers corresponding to each constraint in , and Respectively and The corresponding dual multiplier.
[0100] S3. Initialize the number of iterations and solve the encryption scheduling model system using the interior point method. Specifically:
[0101] S31, using the interior point method to solve the encrypted power system dispatch model to obtain power supply dispatch parameters; for example, setting the initialization iteration number iter_no equal to 0; when performing the first iterative solution, let iter_no increase by 1, and use the interior point method to solve the power system dispatch model established in step S21 to obtain power supply dispatch parameters, including the dispatch plan power output and
[0102] S32, based on the power supply scheduling parameters, the encrypted heating system scheduling model is solved by the interior point method to obtain the optimal objective function value corresponding to the heating system scheduling model; that is, based on the scheduling plan power output and The interior point method is used to solve the heating system scheduling model established in step S24.
[0103] S4: Determine whether the optimal objective function value corresponding to the current encrypted scheduling model system is the preset value. If so, converge to the global optimal solution to obtain the scheduling solution corresponding to the scheduling model system; for example, if the optimal objective function value of the minimum infeasibility model in S24 is equal to 0, then converge to the global optimal solution use Decryption can obtain the optimal collaborative scheduling solution and costs.
[0104] If not, the maximum non-dominated cutting plane inequality is generated according to the encrypted heating system scheduling model and the power supply scheduling parameters, and the encrypted scheduling model system is re-solved according to the maximum non-dominated cutting plane inequality as a constraint condition, that is, returning to step S3 to re-solve the power system scheduling model and the heating system scheduling model; wherein the maximum non-dominated cutting plane inequality includes:
[0105]
[0106] Embodiment 2
[0107] Please refer to Figure 2 A fully privacy-protected energy scheduling device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of a fully privacy-protected energy scheduling method as described in Example 1 is implemented.
[0108] In summary, the fully privacy-protected energy scheduling method and device provided by the present invention combines the encryption method based on elementary transformation of parameter matrices with the decomposition algorithm, and realizes the distributed collaborative scheduling of the power system and the regional heating system under the premise of ensuring the accuracy of the results and the efficiency of calculation; compared with the traditional distributed scheduling method of the electric-thermal coupling system, the model parameters are used to perform elementary transformations on the parameter matrix of the scheduling problem by using random sparse non-singular matrices, which conceals the private information in the parameter matrix, avoids the risk of information leakage caused by semi-honest opponents and external eavesdroppers, and ensures the security of private information; and compared with the traditional encryption method, the present invention greatly improves the distributed solution performance of the algorithm by solving the minimum infeasibility problem to generate the maximum non-dominated cutting plane; when the method in the present invention is applied to the scenario of scheduling plan formulation of the electric-thermal coupling multi-energy flow system, it can be adapted to the original power system and regional heating system energy management system, which is beneficial to reduce operating costs while ensuring information security, and at the same time improve the energy efficiency of the electric-thermal coupling multi-energy flow system.
[0109] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A fully privacy-protected energy scheduling method, characterized in that: include: Obtain the scheduling model system to be encrypted; Performing elementary transformation on the parameter matrix in the scheduling model system by using a random sparse non-singular matrix to obtain an encrypted scheduling model system; Initialize the number of iterations and solve the encryption scheduling model system by an interior point method; Determine whether the optimal objective function value corresponding to the current encrypted scheduling model system is a preset value. If so, converge to the global optimal solution to obtain the scheduling plan corresponding to the scheduling model system.
2. A fully privacy-protected energy scheduling method according to claim 1, characterized in that: The scheduling model system includes a heating system scheduling model and a power system scheduling model; The initializing the number of iterations and solving the encryption scheduling model system by the interior point method includes: The encrypted power system dispatch model is solved by using the interior point method to obtain power supply dispatch parameters; Based on the power supply scheduling parameters, the encrypted heating system scheduling model is solved using the interior point method to obtain the optimal objective function value corresponding to the heating system scheduling model.
3. A fully privacy-protected energy scheduling method according to claim 2, characterized in that: The power system dispatch model includes: Among them, p is the dispatch decision variable in the power system dispatch model, ξ is a non-negative auxiliary variable, is a non-negative slack variable; a and b are the coefficient matrices corresponding to the objective function; E p and F p is the coefficient matrix corresponding to the constraints, U p is the coefficient matrix associated with the slack variables; The power supply system dispatcher generates a random sparse non-singular matrix Y p and I p The elementary transformation of the parameter matrix in the scheduling model system includes: Among them, Y p and I p represents a sparse non-singular matrix; 4. A fully privacy-protected energy scheduling method according to claim 2, characterized in that: The heating system scheduling model includes: Among them, α, h and τ are variable vectors representing the parameters of the heating system scheduling model, s h are the slack variables, c, E, U, G (ψ∈{α, h, τ}) are the corresponding parameter matrices in the objective function and constraints in the heating system scheduling model; The heating system dispatcher generates a random sparse non-singular matrix Y ψ and I ψ (ψ∈{α,h,τ}) performs elementary transformation on the parameter matrix in the scheduling model system, including: in, It is the boundary power information transmitted from the power supply system to the heating system.
5. A fully privacy-protected energy scheduling method according to claim 4, characterized in that: Before solving the encrypted heating system scheduling model based on the power supply scheduling parameters by using the interior point method, the method further includes: The encrypted heating system scheduling model is equivalently expressed as a feasibility problem to obtain a feasibility model; A normalization condition is obtained, and the normalization condition is combined with the dual problem of the feasibility model to obtain a minimum infeasibility problem model.
6. A fully privacy-protected energy scheduling method according to claim 5, characterized in that: The feasibility model includes: in, Boundary cost information transferred from the power supply system to the heating system.
7. A fully privacy-protected energy scheduling method according to claim 6, characterized in that: The normalization conditions include: The minimum infeasibility problem model includes: in, and Respectively The dual multipliers corresponding to each constraint in , and Respectively and The corresponding dual multiplier.
8. A fully privacy-protected energy scheduling method according to claim 7, characterized in that: The determining whether the optimal objective function value corresponding to the current encryption scheduling model system is a preset value also includes: If not, generating a maximum non-dominated cut plane inequality according to the encrypted heating system scheduling model and power supply scheduling parameters; The encrypted scheduling model system is re-solved based on the maximum non-dominated cut plane inequality as a constraint condition.
9. A fully privacy-protected energy scheduling method according to claim 8, characterized in that: The generating of the maximum non-dominated cut plane inequality according to the encrypted heating system scheduling model and power supply scheduling parameters comprises:
10. A fully privacy-protected energy scheduling device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the processor implements each step of a fully privacy-protected energy scheduling method as described in any one of claims 1-9.