Multi-dimensional evaluation and dynamic scheduling method under power distribution network panoramic perception system

Abstract

The application discloses a kind of multi-dimension evaluation and dynamic scheduling method under distribution network panoramic perception system, it is related to distribution network technical field, including: the operation data of distribution network is collected;Determine the objective function and constraint condition of distribution network dynamic scheduling;Determine the operation mode of distributed power supply node, energy storage node and data center node in distribution network;Based on the uncertainty of each operation mode and renewable energy power generation and load prediction, establish distribution network-station area long time scale centralized optimization scheduling model based on stochastic dual dynamic programming, determine and output the first certainty phase in station area energy storage charging and discharging state and flexible load response time period decision;Adjust station area energy storage charging and discharging state and flexible load response time period decision;The optimal scheduling strategy of each flexibility resource in current period is obtained by using optimization algorithm.The application solves the problem that the accuracy of current distribution network dynamic scheduling needs to be improved.

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, and specifically to a multi-dimensional evaluation and dynamic scheduling method under a panoramic perception system for power distribution networks. Background Technology

[0002] A distribution network refers to a power grid that receives electrical energy from the transmission network or regional power plants and distributes it locally or in stages according to voltage to various users through distribution facilities. With the increasing scale of the distribution network and the increase in faults due to grid aging, the safe and stable control capabilities of the distribution network operation have been severely tested, and dispatchers lack effective technical support for daily operations and dynamic adjustments.

[0003] The operation and maintenance of power distribution networks involve multiple disciplines, long cycles, and multiple nodes. Dispatchers are responsible for integrating and processing the power distribution network operation supervision, power outage and restoration management, operation ticket management, dispatch records, and other tasks that exist in different business systems. There are problems such as long operation time during the parallel execution of dispatch instructions, untimely information reporting, and low efficiency of on-site interaction, which leads to the need to improve the accuracy of dynamic dispatching of the current power distribution network. Summary of the Invention

[0004] In view of the above-mentioned shortcomings in the existing technology, the present invention provides a multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of distribution network, which aims to solve the problem that the accuracy of dynamic scheduling of distribution network needs to be improved.

[0005] To achieve the aforementioned objectives, the present invention employs the following technical solution: a multi-dimensional evaluation and dynamic scheduling method under a panoramic perception system for power distribution networks, comprising the following steps: The operation data of the power distribution network is collected, including voltage and power data of distribution nodes, operation data of distributed power generation nodes, operation data of energy storage nodes, and operation data of data center nodes. The objective function and constraints for dynamic scheduling of the distribution network are determined. The constraints include those for the operation of distributed power generation nodes, energy storage nodes, and data center nodes. Based on the operating data, objective function, and constraints of the distribution network, determine the operating modes of distributed power generation nodes, energy storage nodes, and data center nodes in the distribution network; Based on the uncertainties of various operating modes and renewable energy power generation and load forecasting, a long-term centralized optimization scheduling model for distribution network-transformer area based on stochastic dual dynamic programming is established to determine and output the energy storage charging and discharging status and flexible load response period decision of the transformer area in the first deterministic stage. The system receives the energy storage charging and discharging status and flexible load response time period decisions for the receiving area, and adjusts these decisions based on the actual operating conditions of the current period. The optimal scheduling strategy for each flexible resource in the current time period is obtained by using an optimization algorithm, and the optimal scheduling strategy is then passed to the long-term centralized optimization scheduling model of the distribution network and distribution area.

[0006] Furthermore, the operating data of the distributed power node includes the power data of the distributed power node, and the operating data of the energy storage node includes the power data, battery charge state data, and energy storage capacity data of the energy storage node.

[0007] Furthermore, the objective function for the dynamic scheduling of the distribution network is:

[0008] in, The total cost of dynamic dispatching of the distribution network. For the operating costs of energy storage nodes, For the operating costs of data center nodes, Costs resulting from voltage exceeding limits.

[0009] Furthermore, the formula for calculating the operating cost of the energy storage node is as follows:

[0010] in, For the operating time period of the distribution network, For the distribution network operating time set, This is the number of the energy storage node. It is a collection of energy storage nodes. The unit power dispatch cost of energy storage nodes, For the first One energy storage node in Charging state variables during a time period For the first One energy storage node in Discharge state variables during the time period For the first One energy storage node in Charging power during the period For the first One energy storage node in Discharge power during the period This is the scheduling time step.

[0011] Furthermore, the formula for calculating the operating cost of the data center node is as follows:

[0012] in, For the set of power distribution nodes, Numbering of distribution nodes, express Time-of-use electricity pricing for different time periods for Data center node scheduling to the time period The data load that computing resources process for each power distribution node. Indicates in The average power required by a data center node to process a unit of data load during a given time period. To schedule the time step, Indicates the first Each distribution node is The equivalent power consumption due to data processing load during a certain period This indicates the operating cost of the data center node during that time period.

[0013] Furthermore, the formula for calculating the cost of losses caused by the voltage exceeding the limit is as follows:

[0014] in, The duration of voltage exceeding the limit. For voltage exceeding limits, For the scene The probability of occurrence This is a set of scenarios where the node voltage exceeds the upper limit. This is a set of scenarios where the node voltage falls below the lower limit. This represents the cost of loss per unit time and unit voltage exceeding the limit. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node This represents the total number of power distribution nodes.

[0015] Furthermore, the constraints on the operation of the distributed power node include power constraints on the distributed power node, which are as follows:

[0016]

[0017]

[0018] in, The power of the distributed power nodes. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node and All are droop coefficients. This represents the maximum distributed generation capacity of the distributed power generation node at the distribution node. The maximum load power of the distribution node. For the first The rated voltage of each distribution node.

[0019] Furthermore, the constraints on the operation of the energy storage node include: mutual exclusion constraints on charge and discharge states, charging power constraints, discharging power constraints, constraints between charging and discharging power and battery charge state, energy storage capacity constraints at adjacent times, battery charge state constraints, and energy storage capacity constraints.

[0020] Furthermore, the data center node includes multiple front-end servers, which are used to receive and allocate data processing tasks. Based on the voltage, load, and time-of-use pricing information of each power distribution node, the data center node allocates data processing tasks to data center resources corresponding to different power distribution nodes, or transfers deferred data processing tasks to other operating periods, thereby transferring the equivalent power consumption corresponding to the data load between different power distribution nodes and different time periods. The operating data of the data center node includes the data load requirements corresponding to each front-end server, the data load of each power distribution node, and the equivalent power required to process the corresponding data load. The constraints for the operation of the data center nodes include: The sum of all data loads scheduled by each front-end server equals the total data load demand received by that front-end server. The data load allocated to each power distribution node by each front-end server is equal to the sum of the local unprocessed data load of the front-end server at that power distribution node and the data load transferred from the corresponding data center resource space of other power distribution nodes to that power distribution node for processing; The actual data load processed by each power distribution node scheduled by the data center node is equal to the sum of the local data load of that power distribution node and the data load space transferred in, minus the data load space transferred out and the data load transferred to subsequent time periods for processing; where the data load time transfer amount is the data load that can be delayed to subsequent time periods for processing. Furthermore, based on the uncertainties of various operating modes and renewable energy generation and load forecasting, a long-term centralized optimization scheduling model for the distribution network and transformer substations based on stochastic dual dynamic programming is established to determine and output the charging and discharging status of energy storage in the transformer substations and the decision-making process for flexible load response periods within the first deterministic phase. Specifically, this includes: A forward-backward alternating iterative strategy is adopted to solve the long-term centralized optimization scheduling model of distribution network-transformer area. In the forward iteration process, a finite number of scenarios are sampled at each stage through Monte Carlo simulation to solve the decision variables of each stage. The backward iteration generates the optimal cut as an additional constraint condition to gradually approximate the future cost function and make the model meet the convergence condition, so as to obtain the first deterministic stage of transformer area energy storage charging and discharging status and flexible load response period decision.

[0021] The beneficial effects of this invention are: The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network proposed in this invention can improve the accuracy of dynamic scheduling of the distribution network. By determining the operating mode of distributed power generation nodes, energy storage nodes, and data center nodes in the distribution network based on operating data, the objective function and constraints of dynamic scheduling of the distribution network, it can improve the regulation capability, operational stability and operational economy of the distribution network when there is a risk of node voltage exceeding the limit or power imbalance. When a high proportion of distributed power generation nodes are connected to the distribution network and cause the distribution node voltage to exceed the limit, the distribution network can use droop control to adjust the active power output or reactive power support capability of the distributed power generation nodes to achieve distribution network node voltage regulation. Attached Figure Description

[0022] Figure 1 This is a flowchart of a multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of a power distribution network. Detailed Implementation

[0023] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0024] like Figure 1 As shown, this invention proposes a multi-dimensional evaluation and dynamic scheduling method under a panoramic perception system for distribution networks. This method is applied to distribution networks, which include distributed power generation nodes, energy storage nodes, data center nodes, and multiple distribution nodes. This embodiment includes the following steps: S1: Collect the operation data of the power distribution network, including voltage and power data of distribution nodes, operation data of distributed power generation nodes, operation data of energy storage nodes, and operation data of data center nodes; The operating data of the distributed power node includes the power data of the distributed power node, and the operating data of the energy storage node includes the power data, battery charge status data, and energy storage capacity data of the energy storage node.

[0025] The distributed power nodes here include distributed photovoltaic and distributed wind power; the energy storage nodes are mobile energy storage vehicles; the data center nodes include multiple front-end servers, and the data center nodes schedule the power of the data load of the nodes in the distribution network through the front-end servers.

[0026] S2: Determine the objective function and constraints for dynamic scheduling of the distribution network. The constraints include constraints on the operation of distributed power generation nodes, constraints on the operation of energy storage nodes, and constraints on the operation of data center nodes. The loads that can be connected to the distribution nodes of the power distribution network include data loads and conventional power loads. When the power of the power distribution network is unbalanced, the data center electromechanical system can schedule the power of the data loads at each distribution node in the power distribution network to perform spatiotemporal transfer of the power of the data loads, thereby realizing the control of the voltage and power of each distribution node.

[0027] The objective function for the dynamic scheduling of the distribution network is:

[0028] in, The total cost of dynamic dispatching of the distribution network. For the operating costs of energy storage nodes, For the operating costs of data center nodes, Costs resulting from voltage exceeding limits.

[0029] The formula for calculating the operating cost of an energy storage node is as follows:

[0030] in, For the operating time period of the distribution network, For the distribution network operating time set, This is the number of the energy storage node. It is a collection of energy storage nodes. The unit power dispatch cost of energy storage nodes, For the first One energy storage node in Charging state variables during a time period For the first One energy storage node in Discharge state variables during the time period For the first One energy storage node in Charging power during the period For the first One energy storage node in Discharge power during the period This is the scheduling time step.

[0031] The formula for calculating the operating cost of a data center node is as follows:

[0032] in, For the set of power distribution nodes, Numbering of distribution nodes, express Time-of-use electricity pricing for different time periods for Data center node scheduling to the time period The data load that computing resources process for each power distribution node. Indicates in The average power required by a data center node to process a unit of data load during a given time period. To schedule the time step, Indicates the first Each distribution node is The equivalent power consumption due to data processing load during a certain period This indicates the operating cost of the data center node during that time period.

[0033] The formula for calculating the cost of losses caused by voltage exceeding limits is as follows:

[0034] in, The duration of voltage exceeding the limit. For voltage exceeding limits, For the scene The probability of occurrence This is a set of scenarios where the node voltage exceeds the upper limit. This is a set of scenarios where the node voltage falls below the lower limit. This represents the cost of loss per unit time and unit voltage exceeding the limit. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node This represents the total number of power distribution nodes.

[0035] The constraints on the operation of the distributed power nodes include power constraints on the distributed power nodes, which are as follows:

[0036]

[0037]

[0038] in, The power of the distributed power nodes. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node and All are droop coefficients. This represents the maximum distributed generation capacity of the distributed power generation node at the distribution node. The maximum load power of the distribution node. For the first The rated voltage of each distribution node.

[0039] When the voltage of a distribution node exceeds its upper voltage limit, the distributed generation node reduces its active power output through droop regulation or absorbs reactive power through the inverter to alleviate the problem of excessive node voltage. When the voltage of a distribution node is lower than its lower voltage limit, the distributed generation node increases its adjustable active power output or provides reactive power support through droop regulation to raise the node voltage. When the voltage of a distribution node is within the normal voltage fluctuation range including the upper and lower limits, the distributed generation node maintains normal power generation operation.

[0040] Furthermore, the droop regulation power of distributed generation can be expressed as: when hour:

[0041] when hour:

[0042] when hour:

[0043] in, For the first A power distribution node in the scenario Down Distributed power droop regulation power during different time periods; and This is the droop adjustment coefficient; For the first A power distribution node in the scenario Down The actual voltage value during the time period; and The first The upper and lower voltage limits for each distribution node. Positive values ​​indicate that the active power output of distributed generation needs to be reduced, reactive power needs to be absorbed, or power needs to be absorbed by the supporting energy storage. Negative values ​​indicate that the adjustable active power output needs to be increased, reactive power support needs to be provided, or power needs to be released by energy storage to achieve node voltage regulation.

[0044] The constraints on the operation of the energy storage node include: mutual exclusion constraints on charge and discharge states, charging power constraints, discharging power constraints, constraints between charging and discharging power and battery charge state, energy storage capacity constraints at adjacent times, battery charge state constraints, and energy storage capacity constraints.

[0045] The aforementioned operational constraints on energy storage nodes are used to define the feasible operational boundaries of these nodes. Specifically, the mutual exclusion constraint between charge and discharge states prevents the same energy storage node from being in both charging and discharging states simultaneously within the same time period; the charging power and discharging power constraints limit the charging and discharging power of the energy storage node to within its rated power range; the constraint between charging / discharging power and battery charge state describes the changes in the energy storage node's charge capacity at different times; the energy storage capacity constraint between adjacent time points describes the dynamic recursive relationship between energy storage capacity and charging / discharging behavior; and the battery charge state constraint and energy storage capacity constraint ensure that the energy storage node operates within a safe range. These operational constraints on energy storage nodes, along with the operational constraints on distributed power generation nodes and the data load transfer constraints on data center nodes, constitute the constraint system of the distribution network-transformer area collaborative scheduling model.

[0046] Specifically, the operational constraints of energy storage nodes can be expressed as:

[0047]

[0048]

[0049]

[0050]

[0051]

[0052] in, and The first One energy storage node in The charging state variables and discharging state variables for each time period; and The first The maximum charging power and maximum discharging power of each energy storage node; For the first One energy storage node in Energy storage capacity during a given time period For the first One energy storage node in Energy storage capacity for a given time period; For the first One energy storage node in Battery charge state over a period of time; and These are the charging efficiency and the discharging efficiency, respectively.

[0053] The aforementioned constraint system provides a constraint basis for subsequent long-term centralized optimization and short-term rolling adjustment, and dynamically schedules various flexible resources based on the current operating status.

[0054] Specifically, energy storage nodes are flexible adjustment resources in the distribution network. They can perform charging, discharging, or standby operations at different operating times based on the operating status of the distribution area, time-of-use electricity prices, battery charge status, and energy storage capacity constraints. When the energy storage node is a mobile energy storage device, it can also provide power support to different distribution nodes by changing its access location.

[0055] The data center node includes multiple front-end servers, which receive and allocate data processing tasks. Based on the voltage, load, and time-of-use pricing information of each power distribution node, the data center node distributes data processing tasks to the corresponding data center resources of different power distribution nodes, or transfers deferred data processing tasks to other operating periods, thereby transferring the equivalent power consumption corresponding to the data load between different power distribution nodes and different time periods. The operating data of the data center node includes the data load requirements of each front-end server, the data load of each power distribution node, and the equivalent power required to process the corresponding data load.

[0056] The constraints for the operation of the data center nodes include: The sum of all data loads scheduled by each front-end server equals the total data load demand received by that front-end server. The data load allocated to each power distribution node by each front-end server is equal to the sum of the local unprocessed data load of the front-end server at that power distribution node and the data load transferred from the corresponding data center resource space of other power distribution nodes to that power distribution node for processing; The actual data load of each power distribution node scheduled by the data center node is equal to the sum of the local data load of the power distribution node and the data load space transfer in, minus the data load space transfer out and the data load transferred to subsequent time periods for processing; where the data load time transfer amount is the data load that can be delayed to subsequent time periods for processing.

[0057] S3: Based on the operating data, objective function, and constraints of the distribution network, determine the operating modes of distributed power generation nodes, energy storage nodes, and data center nodes in the distribution network; Specifically, the operation modes of the distributed power nodes include normal power generation operation mode, power limiting operation mode, and droop adjustment operation mode; the operation modes of the energy storage nodes include charging operation mode, discharging operation mode, and standby operation mode; the operation modes of the data center nodes include local processing operation mode, data load spatial transfer operation mode, data load time transfer operation mode, and maintaining current data load processing mode.

[0058] Based on the actual voltage value of the distribution node, node power, time-of-use electricity price, predicted output of distributed power sources, battery charge status of energy storage nodes, energy storage capacity, data load demand of data center nodes, as well as objective functions and constraints, the operating mode of each node is determined.

[0059] When the voltage of a distribution node is between the upper and lower voltage limits, the distributed generation node operates in normal power generation mode. When the voltage of a distribution node exceeds the upper voltage limit, the distributed generation node operates in power limiting mode or droop regulation mode to reduce the node voltage. When the voltage of a distribution node is lower than the lower voltage limit, the distributed generation node operates in droop regulation mode to support the node voltage.

[0060] Under the premise of meeting the constraints of energy storage capacity and battery charge state, when the output of distributed power source is high, the electricity price is low, or the node voltage is high, the energy storage node shall prioritize the charging operation mode; when the load is high, the electricity price is high, or the node voltage is low, the energy storage node shall prioritize the discharging operation mode; when the battery charge state of the energy storage node is close to the constraint boundary or the current period does not need to participate in power regulation, the energy storage node shall adopt the standby operation mode.

[0061] When the voltage of a power distribution node is within the allowable range and the load level does not exceed the set threshold, the data center node adopts the local processing operation mode or maintains the current data load processing mode; when the load of a power distribution node is heavy or the voltage is close to the lower limit, the data center node will transfer part of the data load to the data center resources corresponding to other power distribution nodes for processing; when the time-of-use electricity price is high or the power distribution network is under great pressure during a certain period, the data center node will transfer the data load that can be delayed to the subsequent period for processing.

[0062] Specifically, the operational data also includes time-of-use electricity prices for the distribution network.

[0063] S4: Based on the uncertainties of various operating modes and renewable energy generation and load forecasting, a long-term centralized optimization scheduling model for distribution networks and transformer substations based on stochastic dual dynamic programming is established. This model determines and outputs the charging and discharging status of energy storage in the transformer substations and the decision-making process for flexible load response during the first deterministic phase. Specifically, this includes: A forward-backward alternating iterative strategy is adopted to solve the long-term centralized optimization scheduling model of distribution network-transformer area. In the forward iteration process, a finite number of scenarios are sampled at each stage through Monte Carlo simulation to solve the decision variables of each stage. The backward iteration generates the optimal cut as an additional constraint condition to make the model meet the convergence condition, so as to obtain the first deterministic stage of transformer area energy storage charging and discharging status and flexible load response time period decision.

[0064] Specifically, the long-term centralized optimization dispatch model for the distribution network and distribution areas is a multi-stage stochastic optimization model that considers the uncertainties of renewable energy output and load forecasting. The model aims to minimize the expected total operating cost of the distribution network within the dispatching cycle, and the objective function is:

[0065] in, This is a set of renewable energy output and load forecast scenarios generated by Monte Carlo simulations. For the operating scenario, For the scene The probability of occurrence For the set of scheduling time periods, For the scene Down Operating costs of time-of-use energy storage nodes For the scene Down Time-based data center node operating costs For the scene Down Over-limit voltage during certain periods results in cost losses.

[0066] The operating cost of the energy storage node can be expressed as:

[0067] in, It is a collection of energy storage nodes. The unit power dispatch cost of energy storage nodes, and The first Energy storage nodes in the scenario Down Charging and discharging power during the time period and These are the charging state variables and the discharging state variables, respectively.

[0068] The operating cost of the data center node can be expressed as:

[0069] in, For the scene Down Data center node scheduling to the time period The data load that computing resources process for each power distribution node. for Time-of-use electricity pricing for The average power required by a data center node to process a unit of data load during a given time period. This is the scheduling time step.

[0070] To facilitate solving in a multi-stage stochastic optimization model, the voltage over-limit loss can also be equivalently represented by the node voltage over-limit amount and the node voltage under-limit amount.

[0071] The cost of voltage over-limit loss can be expressed as:

[0072] in, and The first A power distribution node in the scenario Down The voltage range is from the upper limit to the lower limit during a given time period. This represents the loss cost corresponding to a unit voltage exceeding the limit per unit time.

[0073] Voltage limits must be met:

[0074]

[0075]

[0076] The decision variables of the model include the charging state variables, discharging state variables, charging power and discharging power of the energy storage node, the droop adjustment power of the distributed power node, the spatial transfer amount of the data load and the temporal transfer amount of the data center node, and the flexible load response period.

[0077] The constraints of the model include distributed power node operation constraints, energy storage node charging and discharging constraints, data center node data load transfer constraints, power distribution node power balance constraints, and power distribution node voltage constraints.

[0078] The power balance constraint at the distribution node can be expressed as:

[0079] in, For the first A power distribution node in the scenario Down Distributed power output during different time periods In order to exchange power with the upper-level power grid, For normal load power, This represents the equivalent load power of the data center. and These represent the charging power and discharging power of the energy storage corresponding to the power distribution node, respectively.

[0080] The equivalent load power of a data center can be expressed as:

[0081] in, For the first A power distribution node in the scenario Down The equivalent load power of the data center due to data processing load during a given period. For the scene Down Time slot scheduling to the The data load processed by each power distribution node Average power per unit of data load.

[0082] The voltage constraint at the distribution node can be expressed as:

[0083] Or, when short-term overruns are permitted and overrun loss costs are included, it can be expressed as:

[0084] in, For the first A power distribution node in the scenario Down The actual voltage value during the time period, and The first The upper and lower voltage limits of each distribution node.

[0085] The data load time shift constraint of a data center node can be expressed as:

[0086] in, Indicates the first A front-end server in Time slots allocated to the The data load processed by each power distribution node Indicates the first A front-end server in Data load requirements received during the time period This indicates the amount of data load that will be delayed until a later time period.

[0087] The spatial transfer constraint of data load in a data center node can be expressed as:

[0088] in, Indicates the first A power distribution node in the scenario Down The actual data load processed during the time period Indicates local data load. Indicates from the first The power distribution node was transferred to the first... The data load processed by each power distribution node Indicates from the first The power distribution node was transferred to the first... The amount of data load processed by each power distribution node.

[0089] The data load processing capacity constraint of a data center node can be expressed as:

[0090] in, For the first Each power distribution node corresponds to the upper limit of the data load that the data center resources can handle in a single time period.

[0091] When using stochastic dual dynamic programming to solve the problem, in the forward iteration process, the decision variables of each stage are solved based on a finite number of running scenarios generated by Monte Carlo simulation; in the backward iteration process, the optimal cut is generated based on the dual information of the subproblems of each stage, and the optimal cut is added to the preceding stage model as an approximate constraint of the future cost function until the convergence condition is met.

[0092] For the energy storage charging and discharging state variables and the flexible load response state variables, a linearization or phased fixed strategy is adopted to ensure that the sub-problems of each stage meet the requirements of stochastic dual dynamic programming.

[0093] The first deterministic phase is the first runtime segment that needs to be actually executed within the current rolling scheduling cycle. The output results include the charging, discharging or standby status of the energy storage nodes in the distribution area, the response time of flexible loads, the spatial and temporal transfer decisions of data loads of data center nodes, and the droop adjustment power of distributed power nodes.

[0094] S5: Receive the energy storage charging and discharging status and flexible load response time period decision of the distribution area, and adjust the energy storage charging and discharging status and flexible load response time period decision of the distribution area according to the actual operation of the current period; S6: Use optimization algorithms to solve for the optimal scheduling strategy of each flexible resource in the current time period, and pass the optimal scheduling strategy to the long-term centralized optimization scheduling model of distribution network-transformer area.

[0095] Specifically, based on the principle of minimizing the operating cost of the distribution network, a centralized optimization scheduling model is used to determine the scheduling decisions for diverse heterogeneous resources. These decisions are then transmitted to the distributed scheduling stage, thereby achieving effective control over the operating cost of the distribution network.

[0096] The fast distributed scheduling method is used to solve the model, which not only gives full play to the advantages of the autonomous distribution area, but also optimizes the control strategy of diverse and heterogeneous flexible resources, thereby improving the operating efficiency and reliability of the distribution network.

[0097] Furthermore, this invention also considers the uncertainties in renewable energy generation and load forecasting, and solves the problems faced by traditional scheduling methods by constructing a long-term centralized optimization scheduling model and a short-term distributed rolling optimization scheduling model.

[0098] The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network proposed in this invention can improve the accuracy of dynamic scheduling of the distribution network. By determining the operating mode of distributed power nodes, energy storage nodes, and data center nodes in the distribution network based on operating data, the objective function and constraints of dynamic scheduling of the distribution network, it can improve the recovery capability of the distribution network and enhance the operation stability and economy of the distribution network under the condition of voltage and power imbalance. When a high proportion of distributed power nodes are connected to the distribution network and cause the voltage of the distribution nodes in the distribution network to exceed the limit, the distribution network can use droop control to adjust the power absorbed or generated by the distributed power nodes to achieve voltage regulation of the distribution network.

[0099] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of the invention.

Claims

1. A multi-dimensional evaluation and dynamic scheduling method under a panoramic perception system for power distribution networks, characterized in that, Includes the following steps: The operation data of the power distribution network is collected, including voltage and power data of distribution nodes, operation data of distributed power generation nodes, operation data of energy storage nodes, and operation data of data center nodes. The objective function and constraints for dynamic scheduling of the distribution network are determined. The constraints include those for the operation of distributed power generation nodes, energy storage nodes, and data center nodes. Based on the operating data, objective function, and constraints of the distribution network, determine the operating modes of distributed power generation nodes, energy storage nodes, and data center nodes in the distribution network; Based on the uncertainties of various operating modes and renewable energy power generation and load forecasting, a long-term centralized optimization scheduling model for distribution network-transformer area based on stochastic dual dynamic programming is established to determine and output the energy storage charging and discharging status and flexible load response period decision of the transformer area in the first deterministic stage. The system receives the energy storage charging and discharging status and flexible load response time period decisions for the receiving area, and adjusts these decisions based on the actual operating conditions of the current period. The optimal scheduling strategy for each flexible resource in the current time period is obtained by using an optimization algorithm, and the optimal scheduling strategy is then passed to the long-term centralized optimization scheduling model of the distribution network and distribution area.

2. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, The operating data of the distributed power node includes the power data of the distributed power node, and the operating data of the energy storage node includes the power data, battery charge status data, and energy storage capacity data of the energy storage node.

3. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, The objective function for the dynamic scheduling of the distribution network is: in, The total cost of dynamic dispatching of the distribution network. For the operating costs of energy storage nodes, For the operating costs of data center nodes, Costs resulting from voltage exceeding limits.

4. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 3, characterized in that, The formula for calculating the operating cost of the energy storage node is as follows: in, For the operating time period of the distribution network, For the distribution network operating time set, This is the number of the energy storage node. It is a collection of energy storage nodes. The unit power dispatch cost of energy storage nodes, For the first One energy storage node in Charging state variables during a time period For the first One energy storage node in Discharge state variables during the time period For the first One energy storage node in Charging power during the period For the first One energy storage node in Discharge power during the period This is the scheduling time step.

5. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 3, characterized in that, The formula for calculating the operating cost of the data center node is as follows: in, For the set of power distribution nodes, Numbering of distribution nodes, express Time-of-use electricity pricing for different time periods for Data center node scheduling to the time period The data load that computing resources process for each power distribution node. Indicates in The average power required by a data center node to process a unit of data load during a given time period. To schedule the time step, Indicates the first Each distribution node is The equivalent power consumption due to data processing load during a certain period This indicates the operating cost of the data center node during that time period.

6. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 3, characterized in that, The formula for calculating the cost of losses caused by voltage exceeding the limit is as follows: in, The duration of voltage exceeding the limit. For voltage exceeding limits, For the scene The probability of occurrence This is a set of scenarios where the node voltage exceeds the upper limit. This is a set of scenarios where the node voltage falls below the lower limit. This represents the cost of loss per unit time and unit voltage exceeding the limit. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node This represents the total number of power distribution nodes.

7. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, The constraints on the operation of the distributed power nodes include power constraints on the distributed power nodes, which are as follows: in, The power of the distributed power nodes. For the first The actual voltage value of each distribution node. For the first The upper limit of voltage at each distribution node For the first The lower voltage limit of each distribution node and All are droop coefficients. This represents the maximum distributed generation capacity of the distributed power generation node at the distribution node. The maximum load power of the distribution node. For the first The rated voltage of each distribution node.

8. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, The constraints on the operation of the energy storage node include: mutual exclusion constraints on charge and discharge states, charging power constraints, discharging power constraints, constraints between charging and discharging power and battery charge state, energy storage capacity constraints at adjacent times, battery charge state constraints, and energy storage capacity constraints.

9. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, The data center node includes multiple front-end servers, which are used to receive and allocate data processing tasks. Based on the voltage, load, and time-of-use pricing information of each power distribution node, the data center node allocates data processing tasks to data center resources corresponding to different power distribution nodes, or transfers deferred data processing tasks to other operating periods, thereby transferring the equivalent power consumption corresponding to the data load between different power distribution nodes and different time periods. The operating data of the data center node includes the data load requirements of each front-end server, the data load of each power distribution node, and the equivalent power required to process the corresponding data load. The constraints for the operation of the data center nodes include: The sum of all data loads scheduled by each front-end server equals the total data load demand received by that front-end server. The data load allocated to each power distribution node by each front-end server is equal to the sum of the local unprocessed data load of the front-end server at that power distribution node and the data load transferred from the corresponding data center resource space of other power distribution nodes to that power distribution node for processing; The actual data load of each power distribution node scheduled by the data center node is equal to the sum of the local data load of the power distribution node and the data load space transfer in, minus the data load space transfer out and the data load transferred to subsequent time periods for processing; where the data load time transfer amount is the data load that can be delayed to subsequent time periods for processing.

10. The multi-dimensional evaluation and dynamic scheduling method under the panoramic perception system of the distribution network according to claim 1, characterized in that, Based on the uncertainties of various operating modes and renewable energy generation and load forecasting, a long-term centralized optimization scheduling model for distribution networks and transformer substations based on stochastic dual dynamic programming is established. This model determines and outputs the charging and discharging status of energy storage in the transformer substations and the decision-making process for flexible load response during the first deterministic phase. Specifically, this includes: A forward-backward alternating iterative strategy is adopted to solve the long-term centralized optimization scheduling model of distribution network-transformer area. In the forward iteration process, a finite number of scenarios are sampled at each stage through Monte Carlo simulation to solve the decision variables of each stage. The backward iteration generates the optimal cut as an additional constraint condition to gradually approximate the future cost function and make the model meet the convergence condition, so as to obtain the first deterministic stage of transformer area energy storage charging and discharging status and flexible load response period decision.

Images