Power distribution system on-line multi-objective optimization method considering carbon emission cost

Abstract

The invention provides a power distribution system on-line multi-objective optimization method considering carbon emission cost, and the method mainly comprises the steps: 1), constructing an operation cost model for a system comprising a heat production unit, an electricity production unit and a thermoelectric mixed production unit, and enabling optimization objectives to cover heat production, electricity production, carbon emission and operation and maintenance cost, and to meet heat / electricity supply and demand balance and production capacity constraint; (2) an electric energy communication topology is formed for communication links between the power generation units and communication links between the heat generation units and the heat and electricity mixed generation units, and a heat energy communication topology is formed for communication links between the heat generation units and the heat and electricity mixed generation units; 3) based on the constructed cost model, introducing a Lagrange dual coefficient to realize localization processing of the heat / electricity supply and demand balance constraint; 4) based on the electric energy unit communication topology, utilizing a distributed gradient descent theory to carry out electric energy collaborative scheduling, and obtaining a local power generation optimal solution of each unit; and 5) based on the heat energy unit communication topology, formulating a heat energy collaborative scheduling strategy, and giving out a local heat production optimal solution of each unit.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution system operation optimization, and particularly relates to an online multi-objective optimization method for a distribution system considering carbon emission costs. Background Art

[0002] In recent years, with the increasingly severe global climate change problem, reducing carbon emissions has become one of the common goals of the international community. Against this background, the power industry, as a major energy-consuming sector and carbon emission source, its emission reduction work has become crucial. As an important part of the power system, the distribution system needs to consider how to reduce the carbon emission cost during operation while ensuring power supply reliability.

[0003] Traditional distribution system operation optimization methods mainly focus on improving the economy and reliability of the system, and rarely consider the cost factor of carbon emissions. With the wide application of distributed energy sources (such as solar energy, wind energy, etc.) and the increase in electric vehicle charging facilities, the operation of the distribution system faces more uncertainties and complexities. These changes not only affect the power supply-demand balance but also have a direct impact on the carbon emission cost of the system. Most existing distribution system operation optimization methods adopt offline optimization strategies, that is, presetting and optimizing the operation mode for a future period according to historical data. However, this method often fails to make the optimal adjustment in the face of real-time changing load demands. Therefore, there is an urgent need to develop an online optimization method that can respond to system state changes in real time and consider carbon emission costs.

[0004] Document 1, "Collaborative Scheduling Optimization Strategy for Smart Park Distribution Network Considering Carbon Emissions" (Power Supply & Utilization, 2024, Vol. 40, No. 10, pp. 73-80) considers the power interaction between the flow shop and electric vehicle charging stations in a smart park distribution network containing carbon emissions, and proposes a collaborative scheduling optimization strategy for the smart park distribution network based on the information gap decision theory, which can reduce the operation cost and carbon emissions of the system according to the risk preference of the decision maker. However, the method mentioned in this document does not consider the real-time changing load demands of the distribution system and has limitations in scalability. In addition, Document 2, "Low-Carbon Economic Scheduling of Multi-Park Interconnected Distribution System Based on Cooperative Game" (Shandong Electric Power Technology, 2024, Vol. 51, No. 5, pp. 19-29) proposes a low-carbon economic scheduling method based on a stepped carbon trading mechanism for multiple park-level distribution systems, which can achieve the management of the total carbon emissions of the multi-park interconnected system. However, the method mentioned in this document is only applicable to the scenario with a fixed carbon trading price and does not consider the real-time carbon emission price coefficient in the carbon trading market.

[0005] Therefore, it is necessary to develop a method that can not only process the real-time changing electricity load and heat demand of the distribution system in a timely manner, but also perform multi-objective optimization on the premise of considering the carbon emission costs of electricity use and heat use in a distribution system including heat production units, power generation units, and combined heat and power units, so as to reduce the overall operating cost of the integrated energy network. Summary of the Invention

[0006] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide an online multi-objective optimization method for a distribution system considering carbon emission costs. By constructing an electrical energy unit communication topology and a thermal energy unit communication topology for a distribution system including heat production units, power generation units, and combined heat and power equipment, a distributed gradient descent theory-based energy collaborative scheduling optimization method for the heat production units, power generation units, and combined heat and power units in the electrical energy unit communication topology and the thermal energy unit communication topology is designed respectively. On the premise of considering the carbon emission costs of electricity use and heat use in the distribution system, the local generation / heat optimization solutions of all units are given, reducing the overall operating cost of the distribution network including carbon emission costs.

[0007] To achieve the above object, the present invention adopts the following technical solutions: An online multi-objective optimization method for a distribution system considering carbon emission costs, comprising the following steps:

[0008] S1: Establish a distribution system operating cost model considering carbon emission costs for a distribution system including heat production units, power generation units, and combined heat and power units. The optimization cost function includes heat production cost, power generation cost, carbon emission cost, and operation and maintenance cost, and needs to satisfy heat / electricity supply-demand balance constraints, power generation capacity constraints, heat production capacity constraints, power generation ramp constraints, and heat production ramp constraints;

[0009] S2: Construct an electrical energy unit communication topology for a distribution system including heat production units, power generation units, and combined heat and power equipment. The electrical energy unit communication topology is composed of the inherent communication links between the power generation units and the combined heat and power units;

[0010] S3: Construct a thermal energy unit communication topology for a distribution system including heat production units, power generation units, and combined heat and power equipment. The thermal energy unit communication topology is composed of the inherent communication links between the heat production units and the combined heat and power units;

[0011] S4: Establish a Lagrangian dual optimization model corresponding to the distribution system operating cost model including heat production cost, power generation cost, carbon emission cost, and operation and maintenance cost. By introducing two Lagrangian dual coefficients and localizing the global heat / electricity demand, the heat / electricity supply-demand balance constraints are eliminated;

[0012] S5: For the obtained Lagrangian constraint optimization model, based on the power unit communication topology composed of power generation units and combined heat and power units, a power collaborative scheduling optimization method for the distribution system based on the distributed gradient descent theory was constructed, and the local power generation optimization solutions for all power generation units and combined heat and power units were given;

[0013] S6: For the obtained Lagrangian constraint optimization model, based on the heat unit communication topology composed of heat generation units and combined heat and power units, a heat collaborative scheduling optimization method for the distribution system was constructed, and the local heat generation optimization solutions for all heat generation units and combined heat and power units were given.

[0014] As a preferred technical solution of the present invention: The present invention establishes an operating cost model for a distribution system considering carbon emission costs for a distribution system including heat generation units, power generation units, and combined heat and power units, as specifically shown in formula (1):

[0015]

[0016] In formula (1), T is the total number of optimized time periods set according to the application scenario; n is the total number of heat generation units, power generation units, and combined heat and power units included in the distribution system; p i (t) is the power generation of the i-th component unit in the distribution system at time t, and the power generation of the heat generation unit is 0; h i (t) is the heat generation of the i-th component unit in the distribution system at time t, and the heat generation of the power generation unit is 0; represents all the power generation costs of the i-th component unit in the distribution system, including power generation energy consumption costs power generation carbon emission costs and power generation operation and maintenance costs The specific expressions are:

[0017]

[0018] Among them, are the time-varying power generation energy consumption quadratic coefficient, linear coefficient, power generation carbon emission quadratic coefficient, linear coefficient, and operation and maintenance cost coefficient respectively. is all the heat generation costs of the i-th component unit in the distribution system, including heat generation energy consumption costs heat generation carbon emission costs and heat generation operation and maintenance costs The specific expressions are:

[0019]

[0020] Among them, are the time-varying heat generation energy consumption quadratic coefficient, linear coefficient, heat generation carbon emission quadratic coefficient, linear coefficient, and operation and maintenance cost coefficient respectively.

[0021] For the feasible power generation p i (t) and heat production h i (t) of the i-th component unit in the distribution system at time t, the operation cost model of the distribution system considering carbon emission cost established for the distribution system including heat production units, power generation units and combined heat and power production units must satisfy the heat / electricity supply-demand balance constraint, power generation capacity constraint, heat production capacity constraint, power generation ramp constraint and heat production ramp constraint;

[0022] The heat / electricity supply-demand balance constraint of the entire integrated energy network is:

[0023]

[0024] In formula (5), P D (t), H D (t) are the time-varying electricity / heat demands of the entire integrated energy network respectively, and ΔP(t), ΔH(t) are the mismatched electricity / heat energies of the entire integrated energy network respectively;

[0025] The power generation / heat production capacity constraint of the i-th component unit in the distribution system is:

[0026]

[0027] In formula (6), p i , h i are the minimum power generation, maximum power generation, minimum heat production and maximum heat production of the i-th component unit respectively;

[0028] The i-th component unit in the distribution system also needs to satisfy the following power generation / heat production ramp constraints:

[0029]

[0030] In formula (7), is the ramp rate coefficient of the i-th component unit in the distribution system;

[0031] As a preferred technical solution of the present invention: The present invention constructs an electric energy unit communication topology for a distribution system including heat production units, power generation units and combined heat and power production equipment, specifically as follows:

[0032] First, obtain the communication link information of the entire distribution system, screen out all heat production units, and retain all power generation units and combined heat and power production units. At the same time, eliminate all communication links input to or output from heat production units, thereby constructing an electric energy unit communication topology for the distribution system; for the electric energy unit communication topology composed of the inherent communication links between all power generation units and combined heat and power production units, define the corresponding communication weight matrix M(t) = [mij (t):

[0033]

[0034] In formula (8), K i (t) represents the set consisting of all power generation units and combined heat and power units that transmit information to the i-th power generation unit or combined heat and power unit in the power distribution system;

[0035] As a preferred technical solution of the present invention: The present invention constructs a communication topology of thermal energy units for a power distribution system including heat generation units, power generation units, and combined heat and power equipment, specifically as follows:

[0036] First, obtain the communication link information of the entire power distribution system, screen out all power generation units, and retain all heat generation units and combined heat and power units. At the same time, eliminate all communication links input to or output from power generation units, thereby constructing a communication topology of thermal energy units for the power distribution system; for the communication topology of thermal energy units composed of the inherent communication links between all heat generation units and combined heat and power units, define the corresponding communication weight matrix N(t) = [n ij (t):

[0037]

[0038] In formula (9), L i (t) represents the set consisting of all heat generation units and combined heat and power units that transmit information to the i-th heat generation unit or combined heat and power unit in the power distribution system;

[0039] As a preferred technical solution of the present invention: The present invention establishes a Lagrangian dual optimization model corresponding to the operation cost model of a power distribution system including heat generation cost, power generation cost, carbon emission cost, and operation and maintenance cost, and eliminates the heat / electricity supply-demand balance constraint by introducing two Lagrangian dual coefficients and heat / electricity global demand localization, specifically as follows:

[0040] Define the following Lagrangian function:

[0041]

[0042] In formula (10), is the Lagrangian dual variable related to the global electricity supply-demand balance constraint of the power distribution system; is the Lagrangian dual variable related to the global heat supply-demand balance constraint of the power distribution system; is the local electricity demand of the i-th unit in the power distribution system, satisfying It should be noted that the local electricity demand of all heat generation units is 0. is the local heat demand of the $i$-th unit in the power distribution system, satisfying It should be noted that the local heat demands of all power generation units are 0. Based on formula (10), a Lagrangian dual optimization model corresponding to the original power distribution system operation optimization problem can be established:

[0043]

[0044] As a preferred technical solution of the present invention: For the obtained Lagrangian constrained optimization model, based on the power unit communication topology composed of power generation units and combined heat and power units, a power collaborative scheduling optimization method for the power distribution system based on the distributed gradient descent theory is constructed, and the local generation optimization solutions of all power generation units and combined heat and power units are given as follows:

[0045] Taking the $i$-th power generation unit and combined heat and power unit in the power distribution system as an example, the solution steps of the power collaborative scheduling optimization method for the power distribution system based on the distributed gradient descent theory are as follows:

[0046] S6.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of this unit:

[0047]

[0048] where is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology. S6.2: Update the set $K$ i (t) of units that need to transmit optimization information currently.

[0049] S6.3: Based on $K$ i (t) obtained in step S7.2, send the optimization information in the previous update to the units in the set $K$ i (t) of units that need to transmit optimization information currently S6.4: Receive the local optimization information sent by other nodes in the power unit communication topology

[0050] S6.5: Based on what is received in step S6.4 update the local aggregated auxiliary variable and the local weighted auxiliary variable as follows:

[0051]

[0052] S6.6: Based on what is obtained in step S6.5 and update the Lagrangian dual variable related to the global power supply-demand balance constraint of the power distribution system

[0053]

[0054] S6.7: Update the secondary coefficient of power generation energy consumption, the primary coefficient, the secondary coefficient of power generation carbon emissions, the primary coefficient, and the operation and maintenance cost coefficient in real time as well as the local electricity demand of all power generation units and combined heat and power production units in the power unit communication topology S6.8: Based on the Lagrangian dual variables obtained in step S6.6 calculate the local optimized power generation p of all power generation units and combined heat and power production units in the power unit communication topology i (t + 1):

[0055]

[0056] The maximization operation for and the cancellation operation for in formula (14) can ensure that the power generation capacity constraint and the power generation ramp-up constraint are surely satisfied.

[0057] S6.9: Generate an updated optimization step size It can also be generated according to other rules, but the following constraints must be satisfied:

[0058]

[0059] S6.10: Based on the local optimized power generation p of all power generation units and combined heat and power production units in the power unit communication topology calculated in step S6.8 i (t + 1), calculate what is required for the next update process

[0060]

[0061] is based on p obtained in S6.8 i (t + 1) and the updated in S6.7 to calculate the optimization gradient:

[0062]

[0063] S6.11: Determine the condition t < T. If it is satisfied, repeat step S6.2. If it is not satisfied, end all steps.

[0064] As a preferred technical solution of the present invention: For the obtained Lagrangian constraint optimization model, based on the thermal energy unit communication topology composed of heat production units and combined heat and power production units, a thermal energy collaborative scheduling optimization method for the distribution system is constructed, and the local heat production optimization solutions of all heat production units and combined heat and power production units are given as follows:

[0065] Taking the i-th heat production unit and combined heat and power production unit in the distribution system as an example, the solution steps of the thermal energy collaborative scheduling optimization method for the distribution system are as follows:

[0066] S7.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of this unit:

[0067]

[0068] where is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology.

[0069] S7.2: Update the unit set L i (t) that needs to transmit optimization information currently.

[0070] S7.3: Based on L i (t) obtained in step S7.2, send the optimization information in the previous update to the units in the unit set L i (t) that needs to transmit optimization information currently S7.4: Receive the local optimization information sent by other nodes in the thermal energy unit communication topology

[0071] S7.5: Based on what is received in step S7.4 update the local aggregated auxiliary variable and the local weight auxiliary variable by performing the following update operations:

[0072]

[0073] S7.6: Based on what is obtained in step S7.5 and update the Lagrangian dual variable

[0074]

[0075] S7.7: Update the quadratic coefficient of heat production energy consumption, the primary coefficient, the quadratic coefficient of heat production carbon emissions, the primary coefficient and the operation and maintenance cost coefficient in real time and the local heat demand of all heat - generating units and combined heat and power units in the heat - energy unit communication topology S7.8: Based on the Lagrangian dual variables obtained in step S7.6 calculate the local optimized heat production h of all heat - generating units and combined heat and power units in the heat - energy unit communication topology i (t + 1):

[0076]

[0077] The maximization operation for and the cancellation operation for in formula (19) can ensure that the heat production capacity constraint and the heat production ramp - up constraint are surely satisfied.

[0078] S7.9: Generate an updated optimization step size It can also be generated according to other rules, but it must satisfy the following constraints:

[0079]

[0080] S7.10: Based on the local optimized heat production h of all heat - generating units and combined heat and power units in the heat - energy unit communication topology obtained in step S7.8 i (t + 1), calculate what is required for the next update process

[0081]

[0082] is based on h i (t + 1) obtained in S7.8 and the updated in S7.7 to calculate the optimization gradient:

[0083]

[0084] S7.11: Determine the condition t < T. If it is satisfied, repeat step S7.2; if not, end all steps.

[0085] For an on - line multi - objective optimization method of a distribution system considering carbon emission cost described in the present invention, compared with the prior art by adopting the above technical solutions, it has the following technical effects:

[0086] 1. Decouple the power / heat production optimization problem of the distribution system through the separation of the power / heat - energy unit communication topology and the separated elimination facing the heat / electricity supply - demand balance constraint.

[0087] 2. Further consider the carbon emission market price factor to optimize and adjust the power / heat production of the distribution system, and overall reduce the operation cost of the distribution network. Description of the Drawings

[0088] Figure 1 is a schematic diagram of the method flow proposed by the present invention;

[0089] Figure 2 is an iterative flowchart of an online multi-objective optimization method for a distribution system considering carbon emission costs;

[0090] Figure 3 is a physical connection structure diagram of the constructed integrated energy simulation system;

[0091] Figure 4 is an electrical / thermal energy unit communication topology diagram in the constructed integrated energy simulation system;

[0092] Figure 5 is a consistency convergence result diagram of local optimization dual variables of 3 power generation units and 2 combined heat and power production units in the constructed simulation system;

[0093] Figure 6 is a consistency convergence result diagram of local optimization dual variables of 3 heat production units and 2 combined heat and power production units in the constructed simulation system;

[0094] Figure 7 is an evolutionary calculation result diagram of local on-line optimization solutions of power generation of 3 power generation units and 2 combined heat and power production units in the constructed simulation system;

[0095] Figure 8 is an evolutionary calculation result diagram of local on-line optimization solutions of heat production of 3 heat production units and 2 combined heat and power production units in the constructed simulation system;

[0096] Figure 9 is a result diagram of the satisfaction of the power supply-demand balance constraint;

[0097] Figure 10 is a result diagram of the satisfaction of the heat supply-demand balance constraint. Detailed Implementation Manner

[0098] The present invention will be further described below with reference to the drawings.

[0099] The present invention proposes an online multi-objective optimization method for a distribution system considering carbon emission costs, which is divided into 6 steps as shown in Figure 1 and is specifically described as follows:

[0100] S1: Establish an operating cost model for a distribution system considering carbon emission costs for a distribution system that includes a heat generation unit, a power generation unit, and a combined heat and power generation unit. The optimization cost function includes heat generation cost, power generation cost, carbon emission cost, and operation and maintenance cost, and needs to satisfy heat / electricity supply-demand balance constraints, power generation capacity constraints, heat generation capacity constraints, power generation ramping constraints, and heat generation ramping constraints;

[0101] S2: Construct an electrical energy unit communication topology for a distribution system that includes a heat generation unit, a power generation unit, and a combined heat and power generation device. The electrical energy unit communication topology consists of the inherent communication links between the power generation unit and the combined heat and power generation unit;

[0102] S3: Construct a thermal energy unit communication topology for a distribution system that includes a heat generation unit, a power generation unit, and a combined heat and power generation device. The thermal energy unit communication topology consists of the inherent communication links between the heat generation unit and the combined heat and power generation unit;

[0103] S4: Establish a Lagrangian dual optimization model corresponding to the operating cost model of the distribution system that includes heat generation cost, power generation cost, carbon emission cost, and operation and maintenance cost. By introducing two Lagrangian dual coefficients and heat / electricity global demand localization, the elimination of heat / electricity supply-demand balance constraints is achieved;

[0104] S5: For the obtained Lagrangian constrained optimization model, based on the electrical energy unit communication topology composed of the power generation unit and the combined heat and power generation unit, construct an electrical energy collaborative scheduling optimization method for the distribution system, and give the local power generation optimization solutions of all power generation units and combined heat and power generation units;

[0105] S6: For the obtained Lagrangian constrained optimization model, based on the thermal energy unit communication topology composed of the heat generation unit and the combined heat and power generation unit, construct a thermal energy collaborative scheduling optimization method for the distribution system, and give the local heat generation optimization solutions of all heat generation units and combined heat and power generation units.

[0106] The present invention first establishes an operating cost model for a distribution system considering carbon emission costs for a distribution system that includes a heat generation unit, a power generation unit, and a combined heat and power generation unit, as shown in formula (1) specifically:

[0107]

[0108] In formula (1), T is the total number of optimization periods set according to the application scenario; n is the total number of heat generation units, power generation units, and combined heat and power generation units included in the distribution system; p i (t) is the power generation amount of the i-th component unit in the distribution system at time t, and the power generation amount of the heat generation unit is 0; h i (t) is the heat generation amount of the i-th component unit in the distribution system at time t, and the heat generation amount of the power generation unit is 0; Denote all the power generation costs of the \(i\)-th component unit in the distribution system, including the power generation energy consumption cost Power generation carbon emission cost and power generation operation and maintenance cost The specific expression is:

[0109]

[0110] wherein, are the time-varying quadratic coefficient, linear coefficient of power generation energy consumption, quadratic coefficient of power generation carbon emission, linear coefficient and operation and maintenance cost coefficient respectively. Denote all the heat production costs of the \(i\)-th component unit in the distribution system, including the heat production energy consumption cost Heat production carbon emission cost and heat production operation and maintenance cost The specific expression is:

[0111]

[0112] wherein, are the time-varying quadratic coefficient, linear coefficient of heat production energy consumption, quadratic coefficient of heat production carbon emission, linear coefficient and operation and maintenance cost coefficient respectively.

[0113] For the feasible power generation \(p i (t)\) and heat production \(h i (t)\) of the \(i\)-th component unit in the distribution system at time \(t\), the operation cost model of the distribution system considering carbon emission cost established for the distribution system including heat production units, power generation units and combined heat and power production units must satisfy the heat / electricity supply-demand balance constraint, power generation capacity constraint, heat production capacity constraint, power generation ramp constraint and heat production ramp constraint;

[0114] The heat / electricity supply-demand balance constraint of the entire integrated energy network is:

[0115]

[0116] In formula (5), \(P D (t), H D (t)\) are the time-varying electricity / heat demands of the entire integrated energy network respectively, and \(\Delta P(t), \Delta H(t)\) are the mismatched electricity / heat energies of the entire integrated energy network respectively;

[0117] The power generation / heat production capacity constraint of the \(i\)-th component unit in the distribution system is:

[0118] In formula (6), p i , h i They are the minimum power generation, maximum power generation, minimum heat production, and maximum heat production of the i-th component unit respectively;

[0119] The i-th component unit in the power distribution system also needs to satisfy the following power / heat ramp constraints:

[0120]

[0121] In formula (7), is the ramp rate coefficient of the i-th component unit in the power distribution system;

[0122] The present invention constructs an electrical energy unit communication topology for a power distribution system including heat production units, power generation units, and combined heat and power production equipment, specifically as follows:

[0123] First, obtain the communication link information of the entire power distribution system, screen out all heat production units, and retain all power generation units and combined heat and power production units. At the same time, eliminate all communication links input to or output from heat production units, thereby constructing an electrical energy unit communication topology for the power distribution system; for the electrical energy unit communication topology composed of the inherent communication links between all power generation units and combined heat and power production units, define the corresponding communication weight matrix M(t) = [m ij (t)]:

[0124]

[0125] In formula (8), Ki(t) represents the set composed of all power generation units and combined heat and power production units that transmit information to the i-th power generation unit or combined heat and power production unit in the power distribution system;

[0126] The present invention constructs a thermal energy unit communication topology for a power distribution system including heat production units, power generation units, and combined heat and power production equipment, specifically as follows:

[0127] First, obtain the communication link information of the entire power distribution system, screen out all power generation units, and retain all heat production units and combined heat and power production units. At the same time, eliminate all communication links input to or output from power generation units, thereby constructing a thermal energy unit communication topology for the power distribution system; for the thermal energy unit communication topology composed of the inherent communication links between all heat production units and combined heat and power production units, define the corresponding communication weight matrix N(t) = [nij(t)]:

[0128]

[0129] In formula (9), Li(t) represents the set composed of all heat production units and combined heat and power production units that transmit information to the i-th heat production unit or combined heat and power production unit in the power distribution system;

[0130] The present invention establishes a Lagrangian dual optimization model corresponding to the distribution system operation cost model including heat production cost, electricity generation cost, carbon emission cost, and operation and maintenance cost. By introducing two Lagrangian dual coefficients and localizing the global heat / electricity demand, the heat / electricity supply-demand balance constraint is eliminated, as follows:

[0131] Define the following Lagrangian function:

[0132]

[0133] In formula (10), is the Lagrangian dual variable related to the global electricity supply-demand balance constraint of the distribution system; is the Lagrangian dual variable related to the global heat supply-demand balance constraint of the distribution system; is the local electricity demand of the i-th unit in the distribution system, satisfying It should be noted that the local electricity demand of all heat production units is 0. is the local heat demand of the i-th unit in the distribution system, satisfying It should be noted that the local heat demand of all electricity generation units is 0. Based on formula (10), a Lagrangian dual optimization model corresponding to the original distribution system operation optimization problem can be established:

[0134]

[0135] For the obtained Lagrangian constraint optimization model, the present invention constructs a distribution system electric energy collaborative scheduling optimization method based on the distributed gradient descent theory for the electric energy unit communication topology composed of electricity generation units and combined heat and power units, and gives the local generation optimization solutions of all electricity generation units and combined heat and power units, as follows:

[0136] Taking the i-th electricity generation unit and combined heat and power unit in the distribution system as an example, the solution steps of the distribution system electric energy collaborative scheduling optimization method based on the distributed gradient descent theory are as follows:

[0137] S6.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of this unit:

[0138]

[0139] where is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology. S6.2: Update the set K i (t) of the units that need to transmit optimization information currently.

[0140] S6.3: Based on K obtained in step S7.2 i(t), send the optimized information in the previous update to the units in the set K of units that currently need to optimize information transmission i (t) S6.4: Receive the local optimized information sent by other nodes in the power unit communication topology

[0141] S6.5: Based on what is received in step S6.4 Perform the following update operations on the local aggregation auxiliary variable and the local weight auxiliary variable :

[0142]

[0143] S6.6: Based on what is obtained in step S6.5 and update the Lagrangian dual variable related to the global power supply - demand balance constraint of the distribution system

[0144]

[0145] S6.7: Update the quadratic coefficient of power generation energy consumption, the primary coefficient, the quadratic coefficient of power generation carbon emissions, the primary coefficient, and the operation and maintenance cost coefficient in real - time as well as the local power consumption demands of all power generation units and combined heat and power units in the power unit communication topology S6.8: Based on the Lagrangian dual variable obtained in step S6.6 calculate the local optimized power generation p of all power generation units and combined heat and power units in the power unit communication topology i (t + 1):

[0146]

[0147] The max operation for and the cancellation operation for in formula (14) can ensure that the power generation capacity constraint and the power generation ramp - up constraint are surely satisfied.

[0148] S6.9: Generate an updated optimization step size It can also be generated according to other rules, but it must satisfy the following constraints:

[0149]

[0150] S6.10: Based on the local optimized power generation p of all power generation units and combined heat and power units in the power unit communication topology calculated in step S6.8 i (t + 1), calculate what is required for the next update process

[0151]

[0152] is based on the p obtained in S6.8 i (t + 1) and the optimized gradient calculated in S6.7:

[0153]

[0154] S6.11: Determine the condition t < T. If it is satisfied, repeat step S6.2; if not, end all steps.

[0155] For the obtained Lagrangian constrained optimization model, the present invention constructs an optimized method for heat energy collaborative scheduling of a distribution system based on the communication topology of heat energy units composed of heat production units and combined heat and power production units, and gives the local heat production optimization solutions of all heat production units and combined heat and power production units, as follows:

[0156] Taking the i-th heat production unit and combined heat and power production unit in the distribution system as an example, the solution steps of the optimized method for heat energy collaborative scheduling of the distribution system are as follows:

[0157] S7.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of the unit:

[0158]

[0159] where is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology.

[0160] S7.2: Update the set L i (t) of units that need to transmit optimization information currently.

[0161] S7.3: Based on L i (t) obtained in step S7.2, send the optimization information in the previous update to the units in the set L i (t) of units that need to transmit optimization information currently. S7.4: Receive the local optimization information sent by other nodes in the heat energy unit communication topology.

[0162] S7.5: Based on what is received in step S7.4 perform the following update operations on the local aggregated auxiliary variable and the local weighted auxiliary variable :

[0163]

[0164] S7.6: Based on what is obtained in step S7.5 and update the Lagrangian dual variables related to the global heat supply-demand balance constraint of the distribution system

[0165]

[0166] S7.7: Update the secondary coefficient of heat production energy consumption, the primary coefficient, the secondary coefficient of heat production carbon emissions, the primary coefficient, and the operation and maintenance cost coefficient in real time as well as the local heat demand of all heat production units and combined heat and power production units in the heat energy unit communication topology S7.8: Based on the Lagrangian dual variables obtained in step S7.6 calculate the local optimized heat production h of all heat production units and combined heat and power production units in the heat energy unit communication topology i (t + 1):

[0167]

[0168] The max operation for and the cancellation operation for in formula (19) can ensure that the heat production capacity constraint and the heat production ramp constraint are surely satisfied.

[0169] S7.9: Generate an updated optimization step size It can also be generated according to other rules, but the following constraints must be satisfied:

[0170]

[0171] S7.10: Based on the local optimized heat production h of all heat production units and combined heat and power production units in the heat energy unit communication topology calculated in step S7.8 i (t + 1), calculate what is required for the next update process

[0172]

[0173] is based on h i (t + 1) obtained in S7.8 and the updated in S7.7

[0174]

[0175] S7.11: Determine the condition t < T. If it is satisfied, repeat step S7.2. If not, end all steps.

[0176] The following gives a simulation example for an integrated energy simulation system:

[0177] The constructed integrated energy simulation system includes 3 power generation units (nodes 1, 2, 8), 3 heat production units (nodes 9, 11, 14), 2 combined heat and power units (nodes 3, 6), and 6 combined heat and power load units (nodes 4, 5, 7, 10, 12, 13). The physical connection structure of the power generation units, heat production units, combined heat and power units, and combined heat and power load units is as Figure 3 shown. According to the composition categories, the communication topologies of the electrical energy units and heat energy units as shown in Figure 4 are constructed. The communication topology of the electrical energy units consists of nodes 1, 2, 3, 6, 8, and the communication topology of the heat energy units consists of nodes 3, 6, 9, 11, 14. The secondary coefficients of power generation energy consumption, primary coefficients, secondary coefficients of carbon emissions from power generation, primary coefficients, and operation and maintenance cost coefficients of all units in the constructed integrated energy simulation system are selected as shown in Table 1.

[0178] Table 1

[0179]

[0180]

[0181] The secondary coefficients of heat production energy consumption, primary coefficients, secondary coefficients of carbon emissions from heat production, primary coefficients, and operation and maintenance cost coefficients of all units in the constructed integrated energy simulation system are selected as shown in Table 2.

[0182] Table 2

[0183]

[0184]

[0185] The minimum power generation, maximum power generation, power generation ramp rate coefficient, and local electricity demand p i , of all units in the power distribution system are selected as shown in Table 3.

[0186] Table 3

[0187]

[0188] The minimum heat production, maximum heat production, heat production ramp rate coefficient, and local heat demand h i , of all units in the power distribution system are selected as shown in Table 4.

[0189] Table 4

[0190]

[0191] Figure 5 Shows the consistency convergence result diagram of the local optimization dual variables of 3 power generation units and 2 combined heat and power units. Figure 6 Shows the consistency convergence result diagram of the local optimization dual variables of 3 heat production units and 2 combined heat and power units. Figure 7 Shows the evolutionary calculation result diagram of the online optimization solution of the local power generation of 3 power generation units and 2 combined heat and power units. Figure 8 Shows the evolutionary calculation result diagram of the online optimization solution of the local heat production of 3 heat production units and 2 combined heat and power units. Figure 9 Shows the result of the satisfaction of the power supply-demand balance constraint. Figure 10 Shows the result of the satisfaction of the heat supply-demand balance constraint.

[0192] The present invention considers the carbon emission cost of power / heat production in a distribution system including heat production units, power generation units, and combined heat and power units, constructs an online collaborative optimization method for the operation of the distribution system based on distributed online gradient optimization technology, gives the local power generation / heat production optimization solutions of all heat production units, power generation units, and combined heat and power units, and reduces the overall operation cost of the integrated energy network including the carbon emission cost.

[0193] The above are only specific embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An online multi-objective optimization method for a power distribution system considering carbon emission costs, characterized in that: The following steps are involved: S1: For the distribution system including heat generation unit, power generation unit and heat and power mixed generation unit, a distribution system operation cost model considering carbon emission cost is established. The optimization cost function includes heat generation cost, power generation cost, carbon emission cost and operation and maintenance cost. It is necessary to meet the heat / electricity supply-demand balance constraint, power generation capacity constraint, heat generation capacity constraint, power generation ramp constraint and heat generation ramp constraint; S2: Construct an energy unit communication topology for the power distribution system including heat generating units, power generating units and heat and power mixed generation equipment. The energy unit communication topology consists of inherent communication links between power generating units and heat and power mixed generation units. S3: Construct a thermal energy unit communication topology for the power distribution system including heat generating units, power generating units and heat and power mixed production equipment. The thermal energy unit communication topology consists of inherent communication links between heat generating units and heat and power mixed production units. S4: Establish a Lagrangian dual optimization model corresponding to the distribution system operation cost model including heat generation cost, electricity generation cost, carbon emission cost and operation and maintenance cost, and eliminate the heat / electricity supply-demand balance constraint by introducing two Lagrangian dual coefficients and localizing the global demand for heat / electricity; S5: Based on the obtained Lagrangian constrained optimization model, and based on the communication topology of the power units composed of power generation units and heat and power mixed production units, a distribution system power coordinated dispatch optimization method based on distributed gradient descent theory is constructed, and the local power generation optimization solution of all power generation units and heat and power mixed production units is given; S6: According to the obtained Lagrangian constrained optimization model, based on the thermal energy unit communication topology composed of heat generating units and heat and power mixed production units, a distribution system thermal energy coordinated scheduling optimization method is constructed, and the local heat production optimization solution of all heat generating units and heat and power mixed production units is given.

2. According to claim 1, an online multi-objective optimization method for a power distribution system considering carbon emission costs is characterized in that: In step S1, a distribution system operation cost model considering carbon emission costs is established for the distribution system including a heat generating unit, a power generating unit and a heat and power mixed generating unit, as shown in formula (1): In formula (1), T is the total number of optimized time periods set according to the application scenario; n is the total number of heat generating units, power generating units, and heat and power mixed generating units included in the power distribution system; p i (t) is the power generation of the ith component unit in the power distribution system at time t, and the power generation of the heat generating unit is 0; h i (t) is the heat generation of the i-th component unit in the power distribution system at time t, and the heat generation of the power generating unit is 0; Represents all power generation costs of the i-th component unit in the distribution system, including the power generation energy cost Carbon emission costs of electricity generation and power generation operation and maintenance costs The specific expression is: in, They are the time-varying secondary coefficient, primary coefficient of power generation energy consumption, secondary coefficient, primary coefficient of carbon emission from power generation and operation and maintenance cost coefficient; is the total heat generation cost of the i-th component in the power distribution system, including the heat generation energy cost Carbon emission costs of heat production and heat generation operation and maintenance costs The specific expression is: in, They are the time-varying secondary coefficient, primary coefficient of heat production energy consumption, secondary coefficient, primary coefficient of heat production carbon emission and operation and maintenance cost coefficient; The feasible power generation p of the i-th component unit in the distribution system at time t is i (t) and heat generation h i (t) The distribution system including the heat generating unit, the power generating unit and the heat and power mixed generating unit must satisfy the heat / electricity supply-demand balance constraint, the power generating capacity constraint, the heat generating capacity constraint, the power generating ramp constraint and the heat generating ramp constraint to establish the distribution system operation cost model taking into account the carbon emission cost; The heat / electricity supply-demand balance constraint of the entire integrated energy grid is: In formula (5), P D (t), H D (t) are the time-varying electricity / heat demand of the entire integrated energy grid, ΔP(t) and ΔH(t) are the mismatched electricity / heat energy of the entire integrated energy grid; The power generation / heat capacity constraint of the i-th component unit in the distribution system is: In formula (6), are the minimum power generation, maximum power generation, minimum heat generation and maximum heat generation of the i-th component unit respectively; The i-th component unit in the power distribution system also needs to meet the following power generation / heat ramping constraints: In formula (7), is the ramp rate coefficient of the i-th component unit in the distribution system.

3. The online multi-objective optimization method for a power distribution system considering carbon emission costs according to claim 1 is characterized in that: The construction of the power unit communication topology for the power distribution system including the heat generating unit, the power generating unit and the heat and power mixed generation equipment described in step S2 is as follows: First, the communication link information of the entire power distribution system is obtained, all heat-generating units are screened out, and all power-generating units and heat-electric mixed-generation units are retained; at the same time, all communication links input to or output from the heat-generating units are eliminated, thereby constructing the power unit communication topology for the power distribution system; for the power unit communication topology composed of the inherent communication links between all power-generating units and heat-electric mixed-generation units, the corresponding communication weight matrix M(t)=[m ij (t)]: In formula (8), K i (t) represents the set of all power generating units and heat and power combined units that transmit information to the ith power generating unit or heat and power combined unit in the distribution system.

4. The online multi-objective optimization method for a power distribution system considering carbon emission costs according to claim 1 is characterized in that: The construction of the thermal energy unit communication topology for the power distribution system including the heat generating unit, the power generating unit and the heat and power mixed generation equipment described in step S3 is as follows: First, the communication link information of the entire power distribution system is obtained, all power-generating units are screened out, and all heat-generating units and heat-electric mixed-generation units are retained; at the same time, all communication links input to or output from the power-generating units are eliminated, thereby constructing a thermal energy unit communication topology for the power distribution system; for the thermal energy unit communication topology composed of the inherent communication links between all heat-generating units and heat-electric mixed-generation units, the corresponding communication weight matrix N(t)=[n ij (t)]: In formula (9), L i (t) represents the set of all heat generating units and heat and power combined units in the power distribution system that transmit information to the i-th heat generating unit or heat and power combined unit.

5. The online multi-objective optimization method for a power distribution system considering carbon emission costs according to claim 1, characterized in that: The Lagrangian dual optimization model corresponding to the distribution system operation cost model including heat generation cost, electricity generation cost, carbon emission cost and operation and maintenance cost is established in step S4, and the heat / electricity supply-demand balance constraint is eliminated by introducing two Lagrangian dual coefficients and localizing the global demand for heat / electricity, as follows: Define the following Lagrangian function: In formula (10), is the Lagrangian dual variable associated with the global electricity supply-demand balance constraint of the distribution system; is the Lagrangian dual variable associated with the global heat supply-demand balance constraint of the distribution system; is the local power demand of the ith unit in the distribution system, satisfying It should be noted that the local electricity demand of all heat-generating units is 0; is the local heat demand of the ith unit in the distribution system, satisfying It should be noted that the local heat demand of all power generation units is 0. Based on formula (10), a Lagrangian dual optimization model corresponding to the original distribution system operation optimization problem can be established:

6. The online multi-objective optimization method for a power distribution system considering carbon emission costs according to claim 1, characterized in that: For the obtained Lagrangian constraint optimization model described in step S5, based on the power unit communication topology composed of power generation units and combined heat and power units, a power collaborative scheduling optimization method for the distribution system based on the distributed gradient descent theory is constructed, and the local generation optimization solutions of all power generation units and combined heat and power units are given as follows: Taking the i-th power generation unit and combined heat and power unit in the distribution system as an example, the solution steps of the power collaborative scheduling optimization method for the distribution system based on the distributed gradient descent theory are as follows: S6.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of this unit: in It is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology; S6.2: Update the unit set K that currently needs to optimize information transmission i (t); S6.3: Based on the K obtained in step S7.2 i (t), to the unit set K that currently needs to perform optimization information transmission i The unit in (t) sends the optimization information in the last update. S6.4: Receive local optimization information sent from other nodes in the power unit communication topology S6.5: Based on the received For local aggregate auxiliary variables and local weight auxiliary variables Perform the following update operations: S6.6: Based on the data obtained in step S6.5 and Update the Lagrangian dual variables associated with the global electricity supply-demand balance constraints of the distribution system S6.7: Real-time update of power generation energy consumption secondary coefficient, primary coefficient, power generation carbon emission secondary coefficient, primary coefficient and operation and maintenance cost coefficient and the local power requirements of all power generating units and heat and power hybrid units in the power unit communication topology S6.8: Based on the Lagrangian dual variables obtained in step S6.6 Calculate the local optimal power generation p of all power generation units and heat and power hybrid units in the power unit communication topology i (t+1): In formula (14), The maximum operation and orientation The cancellation operation can ensure that the power generation capacity constraint and the power generation ramp constraint are met; S6.9: Generate updated optimization step size It can also be generated according to other rules, but must meet the following constraints: S6.10: Based on the local optimized power generation p of all power generating units and heat and power mixed generating units in the power unit communication topology calculated in step S6.8 i (t+1), calculate the time required for the next update process is based on the p obtained in S6.8 i (t+1) is updated with S6.7 The calculated optimization gradient is: S6.11: Determine the condition t < T. If it is satisfied, repeat step S6.2; if not, end all steps.

7. The online multi-objective optimization method for a power distribution system considering carbon emission costs according to claim 1, characterized in that: For the obtained Lagrangian constraint optimization model described in step S6, based on the heat energy unit communication topology composed of heat generation units and combined heat and power units, a heat energy collaborative scheduling optimization method for the distribution system is constructed, and the local heat generation optimization solutions of all heat generation units and combined heat and power units are given as follows: Taking the i-th heat generation unit and combined heat and power unit in the distribution system as an example, the solution steps of the heat energy collaborative scheduling optimization method for the distribution system are as follows: S7.1: Initialize the optimization variables and auxiliary variables that need to be updated during the local optimization process of this unit: in It is used to estimate and eliminate the optimization imbalance error caused by the time-varying directed communication topology; S7.2: Update the unit set L that currently needs to perform optimized information transmission i (t); S7.3: Based on the L obtained in step S7.2 i (t), to the unit set L that currently needs to perform optimization information transmission i The unit in (t) sends the optimization information in the last update. S7.4: Receive local optimization information sent from other nodes in the thermal energy unit communication topology S7.5: Based on the received For local aggregate auxiliary variables and local weight auxiliary variables Perform the following update operations: S7.6: Based on the data obtained in step S7.5 and Update the Lagrangian dual variables associated with the global heat supply-demand balance constraints of the distribution system S7.7: Real-time update of heat generation energy consumption secondary coefficient, primary coefficient, heat generation carbon emission secondary coefficient, primary coefficient and operation and maintenance cost coefficient and the local heat demand of all heat generating units and heat and power hybrid units in the thermal energy unit communication topology S7.8: Based on the Lagrangian dual variables obtained in step S7.6 Calculate the local optimal heat production h of all heat generating units and heat and power mixed generating units in the heat energy unit communication topology i (t+1): In formula (19), The maximum operation and orientation The cancellation operation can ensure that the heat generation capacity constraint and heat generation ramp constraint are met; S7.9: Generate updated optimization step size It can also be generated according to other rules, but must meet the following constraints: S7.10: Based on the local optimized heat production h of all heat generating units and heat and power mixed generating units in the heat energy unit communication topology calculated in step S7.8 i (t+1), calculate the time required for the next update process is based on the h obtained in S7.8 i (t+1) is updated with S7.7 The calculated optimization gradient is: S7.11: Determine the condition t < T. If it is satisfied, repeat step S7.2; if not, end all steps.

Images