Power system optimization scheduling method based on VSL standby response

Abstract

The invention belongs to the technical field of power system optimization scheduling, and particularly relates to a power system optimization scheduling method based on VSL standby response, which comprises the following steps: S1, constructing a ZIP model; s2, establishing a boundary model of VSL power change based on a ZIP model, wherein the boundary model is used for calculating the VSL adjustable power meeting the voltage safety constraint; s3, taking the VSL power change as a schedulable standby resource, and constructing a scheduling optimization model by taking comprehensive benefit maximization as a target; s4, in the actual operation process of the power system, the conditional expectation of the upper and lower total reserve under the capacity shortage scene is calculated based on uncertainty prediction, and the VSL adjustable power is calculated based on the boundary model; taking the calculated conditional expectation of the upper and lower total reserve and the VSL adjustable power as constraints to be input into the scheduling optimization model, and solving a VSL adjustment scheme; and S5, executing the VSL adjustment scheme. The method can give full play to the potential of the VSL as a schedulable standby resource, and improves the flexibility, economy and safety of a power system under the access of high-proportion renewable energy sources.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization scheduling technology, and particularly relates to a power system optimization scheduling method based on VSL reserve response. Background Technology

[0002] As the penetration rate of renewable energy sources such as wind and solar power in the power system continues to increase, the inherent randomness, volatility, and intermittency of their output pose unprecedented challenges to the active power balance, frequency stability, and dispatch operation of the power system. Unlike traditional controllable power sources such as thermal power and hydropower, the power output of renewable energy is highly dependent on weather conditions and is difficult to predict accurately. This can lead to significant power deviations between power generation and load in the system within a short period of time, i.e., unbalanced active power. With a high proportion of renewable energy integrated into the system, the frequency and magnitude of these deviations increase significantly, placing higher demands on the system's flexibility and regulation capabilities.

[0003] Traditional power system dispatching primarily relies on generation-side reserve capacity (such as spinning reserve and non-spinning reserve) to cope with load fluctuations and generation uncertainties. The dispatch center maintains real-time power balance by reserving a certain proportion of upper and lower reserve units to quickly adjust generation output when the system experiences power shortages or surpluses. However, with the increasing proportion of renewable energy, relying solely on generation-side reserves has gradually revealed its limitations: on the one hand, frequent start-ups and shutdowns or deep peak shaving adversely affect the lifespan and economic efficiency of traditional units; on the other hand, the growth of reserve capacity cannot keep pace with the increasing speed and scale of renewable energy fluctuations, leading to strained system flexibility resources and significantly increased dispatching costs.

[0004] Against this backdrop, fully exploring and utilizing load-side flexibility resources has become an important way to improve system regulation capabilities. Among them, voltage-sensitive loads (VSLs), as a type of load with potential regulation capabilities, have received widespread attention in recent years. VSLs refer to electrical equipment whose active power consumption changes significantly with changes in supply voltage, such as incandescent lamps, motor-driven air conditioners and water pumps, and electric heating devices. By fine-tuning the grid voltage, flexible regulation of the power of these loads can be achieved without interrupting user power supply, thereby participating in the system's active power balance control. Due to their wide distribution, fast response speed, and low regulation cost, VSLs are considered a highly promising "virtual backup" resource.

[0005] Although VSLs possess the physical basis for participating in system regulation, they still face numerous technical challenges in practical dispatch applications. First, the regulation capability of VSLs is constrained by voltage safety limitations; excessive voltage regulation may lead to node voltage exceeding limits or instability. Therefore, their adjustable range must be determined while ensuring power quality and system stability. Second, the response of VSLs is decentralized, uncertain, and dependent on user behavior, making centralized and precise control difficult compared to generator sets. Modeling their aggregated response characteristics and quantifying their available reserve capacity is a key challenge for achieving effective dispatch. Furthermore, existing dispatch models often simply treat VSLs as constant loads or ideally controllable resources, lacking comprehensive modeling of their voltage-power dynamics, safety boundaries, and economic incentive mechanisms, resulting in insufficient feasibility of dispatch schemes in actual operation.

[0006] More complexly, in environments of high uncertainty, the system not only needs to know "how much VSL can be dispatched," but also needs to assess whether the total reserve resources (including those on the generation and load sides) are sufficient to cope with risks in extreme scenarios (such as a sudden drop in wind power leading to capacity shortages). Existing methods mostly use fixed reserve ratios or static assessments, which are difficult to reflect the true risk level of the system and lack dynamic quantification methods for "reserve capacity under capacity shortage conditions," thus restricting the scientific nature and robustness of dispatch decisions.

[0007] Therefore, how to fully leverage the potential of VSL as a dispatchable backup resource and enhance the flexibility, economy, and security of the power system under the high proportion of renewable energy integration has become an urgent problem to be solved. Summary of the Invention

[0008] To address the shortcomings of the existing technologies, this invention provides a power system optimization scheduling method based on VSL (Vehicle Service Provider) backup response, which can fully leverage the potential of VSL as a dispatchable backup resource and improve the flexibility, economy, and security of the power system under high-proportion renewable energy access.

[0009] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0010] A power system optimization scheduling method based on VSL (Vehicle Reserve Response) includes the following steps:

[0011] S1. Construct a ZIP model by treating the distribution network as a ZIP load network to quantify the relationship between voltage regulation and VSL power change;

[0012] S2. Based on the ZIP model and power system voltage security constraints, establish a boundary model for VSL power variation to calculate the adjustment boundary of VSL power variation under voltage security constraints, which serves as the adjustable power of VSL.

[0013] S3. Treat VSL power changes as dispatchable reserve resources and combine the impact of VSL power changes on electricity prices to construct a scheduling optimization model with the goal of maximizing comprehensive benefits.

[0014] S4. During the actual operation of the power system, the conditional expectation of the total upper and lower reserves under the capacity shortage scenario is calculated based on the uncertainty prediction, and the adjustable power of VSL is calculated based on the boundary model of S2. The calculated conditional expectation of the total upper and lower reserves and the adjustable power of VSL are used as the constraint inputs of the scheduling optimization model of S3 to solve the VSL adjustment scheme.

[0015] S5. Execute the VSL adjustment scheme obtained in S4 to realize the closed-loop scheduling control of VSL participating in the system standby.

[0016] Compared with the prior art, the present invention has the following advantages:

[0017] 1. Enhancing the diversity and responsiveness of system flexibility resources. Traditional dispatching methods primarily rely on generator units (such as thermal and gas turbines) for backup, resulting in a limited range of flexibility resources. Adjustment speed is constrained by the unit ramp-up rate, and frequent peak shaving impacts equipment lifespan. This method utilizes voltage-sensitive loads (VSLs) as dispatchable backup resources, fully leveraging their widespread distribution and rapid response to expand system regulation capabilities. In emergency scenarios such as sudden wind power drops or load surges, VSLs can achieve millisecond to second-level power responses through voltage fine-tuning, compensating for generation-side regulation delays and significantly enhancing the system's dynamic ability to cope with sudden imbalances.

[0018] 2. Achieving refined modeling and safe, controllable utilization of VSL regulation potential. Existing technologies often treat VSLs as ideal, controllable loads or ignore their voltage dependence characteristics, leading to regulation commands potentially exceeding actual capacity or triggering voltage exceedance risks. This method accurately characterizes the nonlinear relationship between VSL power and voltage using a ZIP model and establishes a regulation boundary model in conjunction with voltage safety constraints, ensuring the exploitation of maximum adjustable power while guaranteeing power quality and system stability. Compared to methods that roughly estimate or fix the regulation range, this scheme improves the feasibility and safety of VSL participation in dispatch.

[0019] 3. Enhance the risk awareness and robustness of dispatch decisions. Traditional reserve configurations often employ deterministic rules (such as the N-1 criterion) or fixed proportions, making it difficult to adapt to the uncertainties brought about by a high proportion of renewable energy. This method introduces the "conditional expectation of total reserve under capacity shortage scenarios" as a dynamic constraint during the operation phase. It performs quantitative evaluation based on the probability distribution of prediction errors for wind power and load, enabling the dispatch model to possess risk awareness capabilities. Compared to static reserve strategies, this method can more scientifically reflect the actual reserve demand of the system, avoiding insufficient or excessive reserve, and improving the robustness and economy of the dispatch scheme.

[0020] 4. Achieving Coordinated Optimization and Enhanced Overall Benefits of Generation-Load Interaction. Existing dispatch models often isolate the interaction between generation and load, or treat demand response only as an auxiliary means. This method aims to maximize overall benefits by coupling the regulation behavior of the VSL (Vehicle Streaming Service) with the electricity pricing mechanism in its modeling. It comprehensively considers operating costs, reserve costs, user response costs, and system reliability to achieve coordinated optimization of generation and load resources. Compared to single-generation-side dispatch, this method can reduce the overall system operating costs, improve the renewable energy absorption capacity, and enhance user participation through reasonable incentive mechanisms.

[0021] In summary, this method integrates the dynamic response characteristics of voltage-sensitive loads, considers voltage security constraints and system risk levels, and achieves coordinated optimization of generation and load. It can fully leverage the potential of VSL as a dispatchable backup resource and improve the flexibility, economy, and security of the power system under high-proportion renewable energy access.

[0022] Preferably, in S2, a ZIP model based on sensitivity linearization is used to establish a boundary model for the VSL adjustable power under voltage safety constraints; the ZIP model based on sensitivity linearization is:

[0023]

[0024] In the formula, and These represent the active power and reactive power injected by node i at time t, respectively. and These represent the active power and reactive power of the load in its ground state, respectively. Let U be the capacity of the reactive power compensation device at node i at time t. N This is the rated voltage amplitude; Let be the voltage change at bus i at time t; p be the ZIP model coefficients for active power, where p1 represents the proportion of constant impedance component, p2 represents the proportion of constant current component, and p3 represents the proportion of constant power component; q be the ZIP model coefficients for reactive power, where q1 represents the proportion of constant impedance component, q2 represents the proportion of constant current component, and q3 represents the proportion of constant power component.

[0025] This ZIP model, with its linearized sensitivity model, considers the nonlinear characteristics of active and reactive power as a function of voltage, and simplifies calculations through linearization. Constraints ensure that the sum of the total sensitivity coefficients equals 1, guaranteeing the model's physical rationality. Traditional methods often use simplified assumptions or empirical formulas to estimate the regulation capability of the VSL, making it difficult to accurately reflect its dynamic response characteristics. This scheme, through a ZIP model based on sensitivity linearization, can finely characterize the complex relationship between VSL power and voltage changes, providing a more accurate regulation boundary. This not only improves the feasibility of VSL participation in scheduling but also avoids the risk of voltage exceedances due to overestimation or underestimation of regulation capability, enhancing the system's operational reliability.

[0026] Preferably, in S2, the power system voltage safety constraints include:

[0027]

[0028] In the formula, L represents the system voltage stability index; L max The preset limit value; L0 represents the L value of the load in the base state; ΔL represents the change in the L value; U C δ represents the column vector voltage magnitude of the node; C Indicates the phase angle of the node voltage;

[0029] The formula for calculating the L value is:

[0030]

[0031] In the formula, Ω D For the set of load nodes; Ω G For connecting the set of power nodes; L i The L value represents node i; N represents the voltage magnitudes at node i and node j, respectively; D N represents the number of load nodes. G Indicates the number of connected power nodes; I L V L These represent the current and voltage vectors of the load node, respectively; I G V G F represents the current and voltage vectors connecting the power nodes, respectively; ij F is calculated as the admittance matrix LG The elements of the matrix, [F LG ]=-[Y LL ] -1 [Y LG ], Y LL This represents the admittance submatrix between load nodes, used to describe the electrical relationships between load nodes; Y LGThe admittance submatrix between load nodes and generator nodes describes how generator nodes are connected to load nodes; H represents the hybrid matrix parameters, which are composed of other submatrices; Z LL Represents matrix Y LL The inverse matrix of F; LG Each element represents the degree of influence of the voltage change at generator node j on the voltage at load node i; a dimensionless voltage transfer coefficient matrix; K GL This indicates the effect of load node current on generator node current; Y GG This represents the admittance submatrix between generator nodes.

[0032] Traditional methods for assessing voltage stability often rely on simplified assumptions or empirical formulas, making it difficult to accurately reflect the system's true state. This approach, by introducing a voltage stability index L and its change ΔL, enables a more precise assessment of the system's voltage stability under different operating conditions. This not only improves the accuracy of the assessment results but also provides a reliable basis for subsequent scheduling decisions. Furthermore, by establishing a strict voltage safety constraint L≤L... max This scheme maximizes the utilization of various regulatory resources (such as generator sets and VSLs) while ensuring system voltage stability. Compared to methods relying solely on experience, this mathematical model-based constraint condition provides more scientific guidance for dispatching decisions, avoiding system instability caused by voltage exceeding limits. Simultaneously, it allows for flexible adjustment of L... max The value can be adjusted to suit the needs of different operating scenarios, improving scheduling flexibility.

[0033] Preferably, in S2, the objective function of the boundary model for VSL power variation is:

[0034]

[0035] In the formula, Let t be the active power of the load injected into node i at time t; Ω Δ express The feasible range; I B This represents all branches in the system that connect node i and node j.

[0036] With this setup, by maximizing and minimizing the objective function, the model can effectively optimize power allocation under different conditions and nodes.

[0037] Preferably, in S2, the equality constraints of the boundary model for VSL power variation include:

[0038]

[0039] In the formula, ΔPC ΔQ C , Δδ C and ΔU C Inject a vector into the node containing active power, reactive power, node voltage phase angle, and node voltage amplitude changes; U C J represents the column vector voltage magnitude of the node; C is the Jacobian matrix of the power flow equation; and This is the sum of the active power and reactive power output of the generator associated with the corresponding node i; and These are the maximum upslope rate and minimum downslope rate of the unit associated with node i, respectively; This refers to the change in line current. The change in the generator's active power; T is the total number of time periods within the scheduling cycle;

[0040] Inequality constraints include:

[0041]

[0042] In the formula, Φ Δ A set of variables for the ground state, including the node voltages. Phase angle of the node Generator active power Reactive power of generator and the square of the transmission line current value ΔΦ Δ The increment set of corresponding variables is represented in the form of sensitivity analysis; and F Δ These represent the upper and lower boundaries of the corresponding variable sets, respectively; δ C Indicates the phase angle of the node voltage; This represents the transmission line current value. The current amplitude on the branch (line) connecting node i and node j under the baseline operating condition.

[0043] This setup provides: 1. Accurate power variation modeling. Through the Jacobian matrix and a detailed description of generator power variations, the model can accurately capture the complex relationship between node power variations and voltage variations in a power system, providing a solid foundation for voltage stability analysis.

[0044] 2. Effective ramp rate control. Generator power changes are strictly limited by the ramp rate, which helps maintain stable operation of the power system under dynamic conditions and avoids voltage instability caused by excessively rapid power changes.

[0045] 3. Flexible variable range management. The upper and lower bounds of the ground-state variables and their increments ensure the safety of the power system under various operating conditions and prevent system failures caused by variables exceeding safe ranges.

[0046] 4. The application of sensitivity analysis methods enables the quantification of the relationship between line current changes and node voltage and phase angle changes, providing strong support for the optimized scheduling and fault diagnosis of power systems.

[0047] In summary, this method constructs a comprehensive and accurate VSL power variation boundary model by comprehensively considering node power variations, generator power variations, ramp rate limits, variable range management, and sensitivity analysis in the power system. This model not only effectively describes the dynamic behavior of the power system but also provides important theoretical support and practical tools for voltage stability analysis, control, and optimization, significantly improving the operational safety and economy of the power system.

[0048] Preferably, in S3, the linearized continuous location marginal price (CLMP) is used to describe the impact of VSL regulation on the electricity sales price at each node; the objective function of the scheduling optimization model is:

[0049]

[0050] In the formula, F G (·) To provide and reserve the associated active power output costs of the generator; F P (·) represents the cost of system airflow limitation and load reduction; F B (·) represents electricity sales revenue; F V (·) represents the adjustment cost of VSL in the ground state; ρ u ρ d and ρ VSL,d (P C ) represents the load shedding, curtailment cost, and electricity sales price function based on linearized CLMP; [E VSL,d E VSL,u [This refers to the potential for changes in system electricity sales revenue after using VSL as a reserve; E] d ( r sys ), Let r represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; G P is the ramp rate of the generator unit. G P represents the active power of the generator. wd This refers to the active power of the wind turbine generator; r0 These are the upper and lower standby capacities associated with VSL, respectively; P lc P represents the active power reduced by the load. C This refers to the electricity sold.

[0051] With this setup, the dispatch optimization model achieves an optimal balance between economic efficiency and stability in the power system by comprehensively considering factors such as generator output costs, wind curtailment and load reduction costs, electricity sales revenue, VSL adjustment costs, and electricity sales prices. The model not only effectively reduces generation and dispatch costs but also maximizes electricity sales revenue, thereby improving the overall operating efficiency and economic benefits of the power system.

[0052] Preferably, in S3, the equality constraints of the scheduling optimization model include:

[0053]

[0054] In the formula, Let be the electrical conductance (real part) between nodes i and j; Let be the susceptance (imaginary part) between nodes i and j; Let be the voltage crossover difference between nodes i and j at time t; Let i be the predicted active power output of the wind farm at node i at time t. Let be the actual active power generated by the wind farm at node i at time t; Let be the normal active load of node i at time t, which is a fixed value; Let be the normal reactive load of node i at time t; Let be the reducible active power load of node i at time t; Let be the reducible reactive load of node i at time t; Let be the active power output of the conventional generator at node i at time t; Let I be the reactive power output of the conventional generator at node i at time t; W It is the set of all nodes connected to the wind farm in the system.

[0055] Preferably, in S3, the inequality constraints of the scheduling optimization model include:

[0056]

[0057] In the formula, The backup capacity provided for the generator at node i; This represents the change in active power output of generator i between two adjacent scheduling periods; I represents the upper limit of the gradeability of generator i; G This is the set of all conventional generator nodes in the system.

[0058] This configuration, through detailed equality and inequality constraints, enables the scheduling optimization model to achieve precise power balance, flexible power source management, comprehensive node power management, effective variable control, and stable generator output in the power grid. These effects work together to significantly improve the stability and operational efficiency of the power grid, optimize energy utilization, and provide strong support for the scheduling and management of smart grids.

[0059] Preferably, in S4, the following method is used to determine whether the power system has insufficient upper or lower reserves;

[0060] Total reserve capacity required by the power system Total reserve capacity and demand It follows a normal distribution:

[0061]

[0062] Based on the distribution of unbalanced active power, when the actual total reserve capacity configured in the power system... Less than the corresponding Or the actual total reserve capacity Less than the corresponding At that time, the upper or lower reserves will be insufficient.

[0063] Preferably, in S4, when calculating the conditional expectation of total upper and lower reserves under a capacity shortage scenario based on uncertainty prediction, the interval between the maximum possible active power shortage and the total reserve capacity is divided into several segments, and the conditional expectation of total upper and lower reserves under a power system capacity shortage is obtained through numerical integration:

[0064]

[0065] in,

[0066]

[0067] in,

[0068] In the formula, Let N represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; seg The number of segments; and Δr is an intermediate variable for piecewise integration. u and Δr d The segmented integration step sizes for the upper total reserve and the lower total reserve are respectively; f u f d The probability density of the power system's reserve demand is a normal distribution on both sides of the mean point; and These represent the maximum requirements for upper and lower reserves in the power system, respectively.

[0069] This configuration, by subdividing the range between the system's maximum potential active power shortage and total reserve capacity, and employing numerical integration methods, enables a more precise quantification of the total reserve demand under uncertain predictions. This helps the power system rationally allocate and schedule reserve resources when facing capacity shortages, thereby improving system stability and reliability. Attached Figure Description

[0070] To make the objectives, technical solutions, and advantages of the invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0071] Figure 1 This is a flowchart of the method;

[0072] Figure 2 This is a schematic diagram of the voltage amplitude and active power injection at the ground state and the maximum regulation node in Example 2;

[0073] Figure 3 This represents the stable margin of the L-index at each loading node in Example 2;

[0074] Figure 4 This is the probability density curve of the power system reserve demand in Example 2;

[0075] Figure 5 This is a schematic diagram showing the node voltage and node load active power before and after VSL is used as a backup in Example 2.

[0076] Figure 6 This is a schematic diagram of the VSL reserve input, the corresponding probability, and the system CLMP when the upper and lower reserve coefficients of the unit are 0.8 and 0.6, respectively, in Example 2.

[0077] Figure 7 This is a schematic diagram showing the reduction in wind power generation, load reduction, and corresponding probabilities when the unit cost coefficient and the storage reduction cost coefficient are 0.8 and 0.6, respectively, in Example 2. Detailed Implementation

[0078] The following detailed explanation illustrates the specific implementation methods:

[0079] Example 1

[0080] like Figure 1 As shown, this embodiment discloses a power system optimization scheduling method based on VSL (Vehicle Reserve Response), including the following steps:

[0081] S1. Construct a ZIP model by treating the distribution network as a ZIP load network to quantify the relationship between voltage regulation and VSL power change;

[0082] S2. Based on the ZIP model and power system voltage security constraints, establish a boundary model for VSL power variation to calculate the adjustment boundary of VSL power variation that satisfies voltage security constraints, which serves as the adjustable power of VSL.

[0083] The general form of the ZIP model is as follows:

[0084]

[0085] in and These represent the active power and reactive power injected by node i at time t, respectively. and These represent the active power and reactive power of the load in its ground state, respectively. Let U be the capacity of the reactive power compensation device at node i at time t. N q represents the rated voltage amplitude. p represents the ZIP model coefficients for active power, where p1 represents the proportion of the constant impedance component, p2 represents the proportion of the constant current component, and p3 represents the proportion of the constant power component; q represents the ZIP model coefficients for reactive power, where q1 represents the proportion of the constant impedance component, q2 represents the proportion of the constant current component, and q3 represents the proportion of the constant power component.

[0086] Then, based on the transmission and distribution coordinated control system architecture, the reactive power in the transmission and distribution network is used to control the loads in the distribution network through unified regulation of the transmission and distribution network. ZIP coefficients are the coefficients of a load model composed of constant impedance, constant current, and constant power loads. The ZIP coefficient load model is used to represent the power consumed by the load as a function of voltage. Meanwhile, this method considers the stochastic economic dispatch method using VSL as a reserve as a steady-state problem, not involving transient processes. Therefore, the distribution network is regarded as a ZIP load network. These networks are installed on the busbars of the transmission and transformation stations and connected to the transmission network through on-load taps.

[0087] In specific implementation, a ZIP model based on sensitivity linearization is adopted to establish a boundary model for the adjustable power of VSL under voltage safety constraints; the ZIP model based on sensitivity linearization is as follows:

[0088]

[0089] In the formula, and These represent the active power and reactive power injected by node i at time t, respectively. and These represent the active power and reactive power of the load in its ground state, respectively. Let U be the capacity of the reactive power compensation device at node i at time t. N This is the rated voltage amplitude; Let i be the voltage change at bus i at time t;

[0090] Considering computational complexity, a linear ZIP model is applied to the voltage control problem that takes sensitivity into account, quickly calculating the power variation of the VSL. In the subsequent scheduling optimization model, the basic ZIP model can be used to describe the base power value of the bus-injected load under a certain voltage setting.

[0091] In practical implementation, an index L is introduced to describe the voltage stability margin constraint. The formula for calculating the value of L is:

[0092]

[0093] In the formula, Ω D For the set of load nodes; Ω G For connecting the set of power nodes; L i The L value represents node i; N represents the voltage magnitudes at node i and node j, respectively; D N represents the number of load nodes. G Indicates the number of connected power nodes; I L V L These represent the current and voltage vectors of the load node, respectively; I G V G F represents the current and voltage vectors connecting the power nodes, respectively; ij F is calculated as the admittance matrix LG The elements of the matrix, [F LG ]=-[Y LL ] -1 [Y LG ], Y LL The Y represents the admittance sub-matrix between load nodes (describing the electrical relationships between load nodes). LG H represents the admittance submatrix between load nodes and generator nodes (describing how generator nodes are connected to load nodes); H represents the hybrid matrix parameters (composed of other submatrices); Z LL Represents matrix Y LL The inverse matrix of F; LG Each element represents the degree of influence of the voltage change at generator node j on the voltage at load node i; a dimensionless voltage transfer coefficient matrix; K GL This indicates the effect of load node current on generator node current; Y GG This represents the admittance submatrix between generator nodes.

[0094] Power system voltage safety constraints include:

[0095]

[0096] In the formula, L represents the system voltage stability index; L max The preset limit value; L0 represents the L value of the load in the base state; ΔL represents the change in the L value; U C δ represents the column vector voltage magnitude of the node; C Indicates the phase angle of the node voltage;

[0097] Traditional methods for assessing voltage stability often rely on simplified assumptions or empirical formulas, which fail to accurately reflect the true state of the system. This approach, by introducing a voltage stability index L and its variation ΔL, enables a more precise assessment of the system's voltage stability under different operating conditions. This not only improves the accuracy of the assessment results but also provides a reliable basis for subsequent scheduling decisions. Furthermore, by establishing a strict voltage safety constraint L≤L... max This scheme maximizes the utilization of various regulatory resources (such as generator sets and VSLs) while ensuring system voltage stability. Compared to methods relying solely on experience, this mathematical model-based constraint condition provides more scientific guidance for dispatching decisions, avoiding system instability caused by voltage exceeding limits. Simultaneously, it allows for flexible adjustment of L... max The value can be adjusted to suit the needs of different operating scenarios, improving scheduling flexibility.

[0098] When using voltage control to regulate the VSL, the load regulation range of a single node is usually limited by the upper and lower limits of the node voltage, resulting in a relatively small regulation range. To simplify the complexity of the model, this method adopts a voltage control scheme based on the current power system operating point sensitivity, thereby obtaining the upper and lower limits of active power regulation based on the current power system VSL.

[0099] The objective function of the boundary model for VSL power variation is:

[0100]

[0101] In the formula, Let t be the active power of the load injected into node i at time t; Ω Δ express The feasible range; I B This represents all branches in the system that connect node i and node j.

[0102] The equality constraints of the boundary model for VSL power variation include:

[0103]

[0104] In the formula, ΔP C ΔQ C , Δδ C and ΔU CInject a vector into the node containing active power, reactive power, node voltage phase angle, and node voltage amplitude changes; U C J represents the column vector voltage magnitude of the node; C is the Jacobian matrix of the power flow equation; and This is the sum of the active power and reactive power output of the generator associated with the corresponding node i; and These are the maximum upslope rate and minimum downslope rate of the unit associated with node i, respectively; This refers to the change in line current. The change in the generator's active power; T is the total number of time periods within the scheduling cycle;

[0105] Inequality constraints include:

[0106]

[0107] In the formula, Φ Δ A set of variables for the ground state, including the node voltages. Phase angle of the node Generator active power Reactive power of generator and the square of the transmission line current value ΔΦ Δ The increment set of corresponding variables is represented in the form of sensitivity analysis; and F Δ These represent the upper and lower boundaries of the corresponding variable sets, respectively; δ C Indicates the phase angle of the node voltage; This represents the transmission line current value. The current amplitude on the branch (line) connecting node i and node j under the baseline operating condition.

[0108] Thus, through the Jacobian matrix and a detailed description of generator power changes, this model can accurately capture the complex relationship between node power changes and voltage changes in a power system, providing a solid foundation for voltage stability analysis. Furthermore, generator power changes are strictly limited by the ramp rate, which helps maintain stable operation of the power system under dynamic conditions and avoids voltage instability caused by excessively rapid power changes. Moreover, the upper and lower bounds of the ground-state variables and their increments ensure the safety of the power system under various operating conditions, preventing system failures caused by variables exceeding safe limits. In addition, the application of sensitivity analysis methods enables the quantification of the relationship between line current changes and node voltage and phase angle changes, providing strong support for optimized power system scheduling and fault diagnosis.

[0109] S3. Treat VSL power changes as dispatchable reserve resources and combine the impact of VSL power changes on electricity prices to construct a scheduling optimization model with the goal of maximizing comprehensive benefits.

[0110] In practice, when VSL (Variable Storage and Licensing) is used as a future reserve investment, it will inevitably affect future electricity sales revenue. Therefore, this method introduces the Continuous Location Marginal Price (CLMP) to describe the impact of VSL regulation on electricity sales prices at each node. Since VSL adjustments may affect the Local Marginal Price (LMP), a linearized CLMP is used to describe the relationship between load value and sales price. In fact, the price that users are willing to pay does not increase with the increase of LMP. Furthermore, as user electricity consumption increases, LMP will approach a price ceiling. Therefore, a price ceiling based on users' willingness to pay for a certain load is adopted.

[0111] By utilizing the upward and downward adjustment intervals established by the VSL at a certain operating point in its base state, the reserve capacity provided by the VSL to the power system is determined. Based on the optimal solution to the real-time scheduling problem, the adjustment range of the VSL and a suitable base state operating point are determined.

[0112] Using VSL as a backup to maximize social welfare, the objective function of the scheduling optimization model is:

[0113]

[0114] In the formula, F G (·) To provide and reserve the associated active power output costs of the generator; F P (·) represents the cost of system airflow limitation and load reduction; F B (·) represents electricity sales revenue; F V (·) represents the adjustment cost of VSL in the ground state; ρ u ρ d and ρ VSL,d (P C ) represents the load shedding, curtailment cost, and electricity sales price function based on linearized CLMP; [E VSL,d E VSL,u [This refers to the potential for changes in system electricity sales revenue after using VSL as a reserve; E] d ( r sys ), Let r represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; G P is the ramp rate of the generator unit. G P represents the active power of the generator. wd This refers to the active power of the wind turbine generator; r0 These are the upper and lower standby capacities associated with VSL, respectively; P lcP represents the active power reduced by the load. C This refers to the electricity sold.

[0115] The first term of the objective function is the cost of generators related to providing active output and reserves; the second term is the system curtailment cost; the third and fourth terms are the risk values ​​of curtailment conditions when the active power imbalance exceeds the total system reserve capacity; the fifth term is the electricity sales revenue; the sixth term is the VSL adjustment cost in the base state; and the seventh term is the conditional value of the risk of changes in system electricity sales revenue considering the VSL adjustment cost after it is used as a reserve.

[0116] The equality constraints of the scheduling optimization model include:

[0117]

[0118] In the formula, Let be the electrical conductance (real part) between nodes i and j; Let be the susceptance (imaginary part) between nodes i and j; Let be the voltage crossover difference between nodes i and j at time t; For XX; Let i be the predicted active power output of the wind farm at node i at time t. Let be the actual active power generated by the wind farm at node i at time t; Let be the normal active load of node i at time t, which is a fixed value; Let be the normal reactive load of node i at time t; Let be the reducible active power load of node i at time t; Let be the reducible reactive load of node i at time t; Let be the active power output of the conventional generator at node i at time t; Let I be the reactive power output of the conventional generator at node i at time t; W It is the set of all nodes connected to the wind farm in the system.

[0119] The inequality constraints of the scheduling optimization model include:

[0120]

[0121] In the formula, The backup capacity provided for the generator at node i; This represents the change in active power output of generator i between two adjacent scheduling periods; I represents the upper limit of the gradeability of generator i; G This is the set of all conventional generator nodes in the system.

[0122] In this way, the scheduling optimization model, through detailed equality and inequality constraints, achieves precise power balance, flexible power source management, comprehensive node power management, effective variable control, and stable generator output in the power grid. These effects work together to significantly improve the stability and operational efficiency of the power grid, optimize energy utilization, and provide strong support for the scheduling and management of smart grids.

[0123] S4. During the actual operation of the power system, the expected conditions of the upper and lower total reserves under the capacity shortage scenario are calculated based on the uncertainty prediction, and the adjustable power of VSL is calculated based on the boundary model of S2. The calculated expected conditions of the upper and lower total reserves and the adjustable power of VSL are used as the constraint inputs of the scheduling optimization model of S3 to solve the VSL adjustment scheme.

[0124] The power of VSL at the ground state operating point is calculated using the ZIP model. Based on this, assuming a power deficit, in future wind power generation, the power system will rely on the upper and lower reserves of generator units to achieve active power balance, with the node active power voltage amplitude adjustments being as follows: (positive adjustment) and (Negative adjustment), the active power adjustment range of VSL is:

[0125]

[0126] In the formula, and These represent the upper and lower reserve capacities associated with VSL at node i, respectively; the meaning of this variable is related to the capacity provided by the generator. and They have the same meaning. This indicates the load reduction achieved by VSL through voltage control, which is equivalent to a new generation of electricity. This indicates the load increment that VSL can achieve through voltage control, which is equivalent to reducing the power generation capacity. The active power injected into the load at any given time; p represents the active power of the load in its ground state; p = [p1p2 p3].

[0127] Total reserve capacity required by the power system Total reserve capacity and demand It follows a normal distribution:

[0128]

[0129] Based on the distribution of unbalanced active power, when the actual total reserve capacity configured in the power system... Less than the corresponding Or the actual total reserve capacity Less than the corresponding At that time, the upper or lower reserves will be insufficient.

[0130] Based on the relationship between the total reserve capacity and reserve demand of the power system, the conditional risks related to the upstream and downstream reserve capacity of the power system can be described as follows:

[0131]

[0132] In the formula, and These represent the maximum demand for upper and lower reserves in the power system, i.e., maximum active power surplus and maximum active power deficit; f u f d It represents a normal distribution on both sides of the probability density mean point; and These represent the active power probabilities corresponding to shortage and surplus conditions, respectively.

[0133] Since the cumulative density function (CDF) of the normal distribution lacks a mathematical expression, it can only be solved using numerical integration. Furthermore, the CDF is non-convex, and piecewise linearization requires a large number of integer variables, making mixed-integer nonlinear programming of reserve decisions difficult to solve. Therefore, this method calculates the conditional expectation based on piecewise numerical integration.

[0134] When calculating the conditional expectation of total upper and lower reserves under capacity shortage scenarios based on uncertainty prediction, the interval between the maximum possible active power shortage and the total reserve capacity is divided into several segments, and the conditional expectation of total upper and lower reserves under power system capacity shortage is obtained through numerical integration:

[0135]

[0136] in,

[0137]

[0138] in,

[0139] In the formula, Let N represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; seg The number of segments; and Δr is an intermediate variable for piecewise integration. u and Δr d The segmented integration step sizes for the upper total reserve and the lower total reserve are respectively; f u f d The probability density of the power system's reserve demand is a normal distribution on both sides of the mean point; and These represent the maximum requirements for upper and lower reserves in the power system, respectively.

[0140] This configuration, by subdividing the range between the system's maximum potential active power shortage and total reserve capacity, and employing numerical integration methods, enables a more precise quantification of the total reserve demand under uncertain predictions. This helps the power system rationally allocate and schedule reserve resources when facing capacity shortages, thereby improving system stability and reliability.

[0141] S5. Execute the VSL adjustment scheme obtained in S4 to realize the closed-loop scheduling control of VSL participating in the system standby.

[0142] Traditional dispatching methods primarily rely on generator units (such as thermal and gas turbines) for backup, offering limited flexibility due to the single type of resource. Regulation speed is constrained by the unit's ramp rate, and frequent peak shaving impacts equipment lifespan. This method treats voltage-sensitive loads (VSLs) as dispatchable backup resources, fully leveraging their widespread distribution and rapid response to expand system regulation capabilities. In emergency scenarios such as sudden wind power drops or load surges, VSLs can achieve millisecond-to-second power responses through voltage fine-tuning, compensating for generation-side regulation delays and significantly enhancing the system's dynamic ability to cope with sudden imbalances. Furthermore, existing technologies often treat VSLs as ideal controllable loads or ignore their voltage dependence, potentially leading to regulation commands exceeding actual capacity or triggering voltage exceedance risks. This method accurately characterizes the nonlinear relationship between VSL power and voltage using a ZIP model and establishes a regulation boundary model based on voltage safety constraints, ensuring the extraction of maximum adjustable power while guaranteeing power quality and system stability. Compared to methods that roughly estimate or fix the regulation range, this approach improves the feasibility and safety of VSL participation in dispatching.

[0143] In addition, traditional reserve configurations often employ deterministic rules (such as the N-1 criterion) or fixed proportions, making it difficult to adapt to the uncertainties brought about by a high proportion of renewable energy. This method introduces the "conditional expectation of total reserve under capacity shortage scenarios" as a dynamic constraint during the operation phase. It performs quantitative evaluation based on the probability distribution of prediction errors for wind power and load, enabling the dispatch model to have risk perception capabilities. Compared to static reserve strategies, this method more scientifically reflects the system's actual reserve demand, avoiding insufficient or excessive reserve, and improving the robustness and economy of the dispatch scheme. Furthermore, existing dispatch models often sever the interaction between generation and load, or only use demand response as an auxiliary means. This method aims to "maximize comprehensive benefits," coupling the VSL's adjustment behavior with the electricity pricing mechanism in its modeling. It comprehensively considers operating costs, reserve costs, user response costs, and system reliability to achieve coordinated optimization of generation and load resources. Compared to single-generation-side dispatch, this method can reduce the overall system operating cost, improve the renewable energy absorption capacity, and enhance user participation through reasonable incentive mechanisms.

[0144] This method integrates the dynamic response characteristics of voltage-sensitive loads, considers voltage security constraints and system risk levels, and achieves generation-load coordinated optimization. It can fully leverage the potential of VSL as a dispatchable backup resource and improve the flexibility, economy, and security of the power system under high-proportion renewable energy access.

[0145] Example 2

[0146] To better illustrate the effectiveness of this method, the following example is provided.

[0147] To verify the effectiveness of the model in this method, a wind farm was added to buses 9, 11, and 15 in the improved IEEE RTS-79 system, with base state operating points of 90MW, 73MW, and 33MW, respectively. The newly added nodes and load conditions are shown in Table 1. The voltage of these buses is limited to [0.95, 1.05] pu, the voltage at reference bus 13 is set to 1.00 pu, and the parameters of the ZIP model are set to 0.2, 0.3, and 0.5. For convenience, power consumption is positive. Furthermore, since this method primarily considers the scheduling problem on a real-time 5-minute scale, all on-load tap changers and capacities are considered known. Therefore, the sample tests were conducted in MATLAB R2013b, with a hardware platform configured as a computer with a 2.8GHz quad-core CPU and 16GB of memory.

[0148] Table 1: Newly Added Network Nodes and Corresponding Loads under Rated Voltage

[0149]

[0150] First, we evaluate the VSL adjustment range considering voltage stability constraints. The ground state is the reactive power optimization result that minimizes network losses in the current system. In the evaluation, the limit of the L index related to voltage stability constraints is 0.2. The evaluation results of the VSL adjustment range are given in this example, as shown in Table 2.

[0151] Table 2: VSL Adjustment Range Considering Voltage Stability Constraints

[0152]

[0153] Under the current base state, the maximum VSL of the power system can be increased to 2871.4MW, and the minimum VSL can be reduced to 2745.4MW. The total reduction is 126.0MW, which is close to 5% of the base state total load; therefore, the regulation potential is huge.

[0154] Figure 2 (a) and (b) show the node voltages and active power of the power system after maximum VSL adjustment based on the VSL evaluation range. Figure 2It can be seen that the power system can increase or decrease the power of the voltage level channel (VSL) (power consumption is positive) by increasing or decreasing the voltage of each node, so as to make full use of the flexibility of the VSL. Since the L index is introduced into the VSL evaluation to ensure the voltage stability margin of the system, the impact of voltage stability constraints on VSL regulation is further studied.

[0155] Table 3 shows the range of VSL values ​​under different L index constraints.

[0156] Table 3: Impact of Static Voltage Stability Margin Constraints on VSL Drop Range

[0157]

[0158] As shown in Table 3, the maximum upward adjustment of VSL remains almost unchanged as the L exponent decreases, but the downward adjustment of VSL varies significantly; specifically, the downward adjustment of VSL decreases as the L exponent decreases. From the definition of the exponent L in S2, it is known that the system's L index depends on the L value at each loading node. Therefore, determining the L value at each loading node is as follows... Figure 3 As shown in the table, the L values ​​at nodes 37 and 39 are very large. Furthermore, it can be seen from the table that the loads at these two nodes are the largest of all nodes, which may lead to system voltage instability. Therefore, when the given L index constraint is reduced to the current L index of these two nodes, it will affect the node voltage control and VSL regulation.

[0159] Next, we consider VSL as a backup power system for real-time dispatch. In this part, the maximum demand and standard deviation of upper and lower reserves can be set according to the system's needs. Generally, in the initial example, the standard deviation of upper and lower reserve demands is set to 3% of the total load, and the maximum reserve demand is 9%. Reserve demand information is shown in Table 4 and... Figure 4 As shown.

[0160] Table 4: Probability Distribution Information of Reserve Demand

[0161]

[0162] Assume the cost of providing upper-level reserves from the generator is 80% of the current marginal cost, and the cost of providing lower-level reserves from the generator is 60% of the current marginal cost. A power system capable of providing sufficient power generally needs to meet the power demand of a given load. Therefore, the reserves provided by VSL in this method can only be used after the generator's reserves are exhausted.

[0163] This example has two scenarios:

[0164] 1) Scenario 1: VSL participates in active power dispatching of the power system as a backup;

[0165] 2) Scenario 2: VSL does not participate in the active power dispatch of the power system as a backup, but only changes the current.

[0166] Voltage participates in the current active power dispatch.

[0167] When VSLs are used as reserves, they can effectively supplement the available active reserve capacity in the system, reducing the cost of generators providing backup. As shown in Table 5, using VSLs as reserves increases total revenue by 14.9%. It is worth noting that although electricity sales profits do not change significantly, VSLs can effectively reduce backup costs and the risk of wind power curtailment / load shedding. Furthermore, Table 6 shows that the system gains additional cap space when VSLs are used as reserves.

[0168] Table 5: Economic scheduling results with and without VSL reserves (5 minutes)

[0169]

[0170] Table 6: Power System Reserves and Wind Curtailment Risks

[0171]

[0172] from Figure 5 (a)(b) It can be seen that regardless of whether VSL is used as a standby, the load and voltage amplitude of the power system are very high, and the difference in increasing electricity sales revenue is small. In order to further study the impact of standby costs of generator sets, the standby costs of different generator sets are calculated, as shown in Table 7.

[0173] Table 7

[0174]

[0175] Figure 6 (a) and (b) show the probabilities of obtaining different CLMP and VSL reserve inputs when the generator's upper reserve cost factor is 0.8 and the lower reserve cost factor is 0.6. Figure 6(a) As shown, at the base-state load level, the electricity sales price per MWh is $50, and the revenue per MWh within 5 minutes is $4.20. When the VSL is not used as a reserve, the system obtains a total of 300MW of upper and lower reserves from the generator side, with a total cost of $618.4. The average reserve cost per MWh of the generator unit is approximately half of the current unit price. However, the VSL can also provide a certain reserve at higher power levels, thereby reducing the system reserve cost and further enhancing the system's ability to meet associated demand and maintain VSL power at a higher level. Correspondingly, reducing the VSL level will not only lead to a loss of electricity sales revenue but also reduce the system's CLMP, thereby reducing the electricity sales price and revenue of the part of the system based on VSL inventory. Therefore, in the various scenarios in Table 7, the system will still maintain the VSL at a higher level and mainly obtain upper limit reserves from the VSL.

[0176] In addition, under the ground state, the system can obtain sufficient up and down reserves from the generator set and VSL to cope with possible future active power fluctuations. Figure 7 (a) and (b) represent the wind cutoff and load reduction values ​​and their corresponding probabilities when the upper and lower storage cost coefficients are 0.8 and 0.6, respectively. At the current storage level, the probability of wind cutoff and load reduction is very small, resulting in low risk costs. Therefore, the current VSL control strategy and unit reserve meet the system's storage capacity requirements.

[0177] In summary, when the reserve cost of a generating unit is not high, the system will increase VSL power. The power system can increase electricity sales revenue while simultaneously storing a certain amount of electricity, thereby reducing the unit's reserve cost. Furthermore, since VSL reserves in this method are mainly used to supplement unit reserves, and the probability of needing VSL reserves is very low, this reduces the risk of using VSL as a reserve to some extent, increasing the importance of VSL power in relation to real-time electricity sales revenue. The results also show that VSL, as an auxiliary power source, can provide a certain reserve capacity for the system and reduce system losses in extreme cases.

[0178] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit the technical solutions. Those skilled in the art should understand that any modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A power system optimal dispatching method based on VSL (Vehicle Reserve Response), characterized in that, Includes the following steps: S1. Construct a ZIP model by treating the distribution network as a ZIP load network to quantify the relationship between voltage regulation and VSL power change; S2. Based on the ZIP model and power system voltage security constraints, establish a boundary model for VSL power variation to calculate the adjustment boundary of VSL power variation under voltage security constraints, which serves as the adjustable power of VSL. S3. Treat VSL power changes as dispatchable reserve resources and combine the impact of VSL power changes on electricity prices to construct a scheduling optimization model with the goal of maximizing comprehensive benefits. S4. During the actual operation of the power system, the conditional expectation of the total upper and lower reserves under the capacity shortage scenario is calculated based on the uncertainty prediction, and the adjustable power of VSL is calculated based on the boundary model of S2. The calculated conditional expectation of the total upper and lower reserves and the adjustable power of VSL are used as the constraint inputs of the scheduling optimization model of S3 to solve the VSL adjustment scheme. S5. Execute the VSL adjustment scheme obtained in S4 to realize the closed-loop scheduling control of VSL participating in the system standby.

2. The power system optimal dispatching method based on VSL reserve response as described in claim 1, characterized in that: In S2, a boundary model for the adjustable power of VSL under voltage safety constraints is established using a ZIP model based on sensitivity linearization; the ZIP model based on sensitivity linearization is as follows: In the formula, and These represent the active power and reactive power injected by node i at time t, respectively. and These represent the active power and reactive power of the load in its ground state, respectively. Let U be the capacity of the reactive power compensation device at node i at time t. N This is the rated voltage amplitude; Let be the voltage change at bus i at time t; p be the ZIP model coefficients for active power, where p1 represents the proportion of constant impedance component, p2 represents the proportion of constant current component, and p3 represents the proportion of constant power component; q be the ZIP model coefficients for reactive power, where q1 represents the proportion of constant impedance component, q2 represents the proportion of constant current component, and q3 represents the proportion of constant power component.

3. The power system optimal dispatching method based on VSL reserve response as described in claim 2, characterized in that: In S2, the power system voltage safety constraints include: In the formula, L represents the system voltage stability index; L max The preset limit value; L0 represents the L value of the load in the base state; ΔL represents the change in the L value; U C δ represents the column vector voltage magnitude of the node; C Indicates the phase angle of the node voltage; The formula for calculating the L value is: In the formula, Ω D For the set of load nodes; Ω G For connecting the set of power nodes; L i The L value represents node i; N represents the voltage magnitudes at node i and node j, respectively; D N represents the number of load nodes. G Indicates the number of connected power nodes; I L V L These represent the current and voltage vectors of the load node, respectively; I G V G F represents the current and voltage vectors connecting the power nodes, respectively; ij F is calculated as the admittance matrix LG The elements of the matrix, [F LG ]=-[Y LL ] -1 [Y LG ], Y LL This represents the admittance submatrix between load nodes, used to describe the electrical relationships between load nodes; Y LG The admittance submatrix between load nodes and generator nodes describes how generator nodes are connected to load nodes; H represents the hybrid matrix parameters, which are composed of other submatrices; Z LL Represents matrix Y LL The inverse matrix of F; LG Each element represents the degree of influence of the voltage change at generator node j on the voltage at load node i; a dimensionless voltage transfer coefficient matrix; K GL This indicates the effect of load node current on generator node current; Y GG This represents the admittance submatrix between generator nodes.

4. The power system optimal dispatching method based on VSL reserve response as described in claim 3, characterized in that: In S2, the objective function of the boundary model for VSL power variation is: In the formula, Let t be the active power of the load injected into node i at time t; Ω Δ express The feasible range; I B This represents all branches in the system that connect node i and node j.

5. The power system optimal dispatching method based on VSL reserve response as described in claim 4, characterized in that: In S2, the equality constraints of the boundary model for VSL power variation include: In the formula, ΔP C ΔQ C , Δδ C and ΔU C Inject a vector into the node containing active power, reactive power, node voltage phase angle, and node voltage amplitude changes; U C J represents the column vector voltage magnitude of the node; C is the Jacobian matrix of the power flow equation; and This is the sum of the active power and reactive power output of the generator associated with the corresponding node i; and These are the maximum upslope rate and minimum downslope rate of the unit associated with node i, respectively; This refers to the change in line current. The change in the generator's active power; T is the total number of time periods within the scheduling cycle; Inequality constraints include: In the formula, Φ Δ A set of variables for the ground state, including the node voltages. Phase angle of the node Generator active power Reactive power of generator and the square of the transmission line current value ΔΦ Δ The increment set of corresponding variables is represented in the form of sensitivity analysis; and F Δ These represent the upper and lower boundaries of the corresponding variable sets, respectively; δ C Indicates the phase angle of the node voltage; This represents the transmission line current value. The current amplitude on the branch connecting node i and node j under the baseline operating condition.

6. The power system optimal dispatching method based on VSL reserve response as described in claim 5, characterized in that: In S3, the linearized continuous location marginal price (CLMP) is used to describe the impact of VSL regulation on the electricity sales price at each node; the objective function of the scheduling optimization model is: In the formula, F G (·) To provide and reserve the associated active power output costs of the generator; F P (·) represents the cost of system airflow limitation and load reduction; F B (·) represents electricity sales revenue; F V (·) represents the adjustment cost of VSL in the ground state; ρ u ρ d and ρ VSL,d (P C ) represents the load shedding, curtailment cost, and electricity sales price function based on linearized CLMP; [E VSL,d E VSL,u [This refers to the potential for changes in system electricity sales revenue after using VSL as a reserve; E] d ( r sys ), Let r represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; G P is the ramp rate of the generator unit. G P represents the active power of the generator. wd This refers to the active power of the wind turbine generator; r0 These are the upper and lower standby capacities associated with VSL, respectively; P lc P represents the active power reduced by the load. C This refers to the electricity sold.

7. The power system optimal dispatching method based on VSL reserve response as described in claim 6, characterized in that: In S3, the equality constraints of the scheduling optimization model include: In the formula, The electrical conductance between nodes i and j; The susceptance between nodes i and j; Let be the voltage crossover difference between nodes i and j at time t; Let i be the predicted active power output of the wind farm at node i at time t. Let be the actual active power generated by the wind farm at node i at time t; Let be the normal active load of node i at time t, which is a fixed value; Let be the normal reactive load of node i at time t; Let be the reducible active power load of node i at time t; Let be the reducible reactive load of node i at time t; Let be the active power output of the conventional generator at node i at time t; Let I be the reactive power output of the conventional generator at node i at time t; W It is the set of all nodes connected to the wind farm in the system.

8. The power system optimal dispatching method based on VSL reserve response as described in claim 7, characterized in that: In S3, the inequality constraints of the scheduling optimization model include: In the formula, The backup capacity provided for the generator at node i; This represents the change in active power output of generator i between two adjacent scheduling periods; I represents the upper limit of the gradeability of generator i; G This is the set of all conventional generator nodes in the system.

9. The power system optimal dispatching method based on VSL reserve response as described in claim 1, characterized in that: In S4, the following method is used to determine whether the power system has insufficient upper or lower reserves; Total reserve capacity required by the power system Total reserve capacity and demand It follows a normal distribution: Based on the distribution of unbalanced active power, when the actual total reserve capacity configured in the power system... Less than the corresponding Or the actual total reserve capacity Less than the corresponding At that time, the upper or lower reserves will be insufficient.

10. The power system optimal dispatching method based on VSL reserve response as described in claim 9, characterized in that: In S4, when calculating the conditional expectation of total upper and lower reserves under a capacity shortage scenario based on uncertainty prediction, the interval between the maximum possible active power shortage and the total reserve capacity is divided into several segments, and the conditional expectation of total upper and lower reserves under a power system capacity shortage is obtained through numerical integration: in, in, In the formula, Let N represent the conditional expected values ​​of the upper total reserve and the lower total reserve, respectively, when the power system experiences a capacity shortage; seg The number of segments; and Δr is an intermediate variable for piecewise integration. u and Δr d The segmented integration step sizes for the upper total reserve and the lower total reserve are respectively; f u f d The probability density of the power system's reserve demand is a normal distribution on both sides of the mean point; and These represent the maximum requirements for upper and lower reserves in the power system, respectively.

Images