Intelligent scheduling method for cascade reservoirs

Abstract

The invention discloses a cascade reservoir intelligent scheduling method, and belongs to the technical field of cascade reservoir intelligent scheduling. The method comprises the following steps: S1, establishing a comprehensive database, and periodically capturing dispatching data of each meteorological observatory, hydrological center, power plant and power grid for model calling calculation; s2, establishing an intelligent simulation model, performing medium-and-long-term runoff rolling forecasting and flood forecasting on a faucet reservoir with seasonal adjustment and performance above as a forecasting object according to data of a database number, and performing medium-term and flood forecasting on all power stations; s3, an operator selects a scheduling object and calculates a time span according to actual conditions, adjusts input and constraint conditions, and selects a proper optimization target according to different scene requirements; and S4, the intelligent simulation model outputs a scheduling scheme according to the input information and supports manual intervention. The technical problems that in traditional cascade reservoir dispatching, judgment depends on artificial experience, decision making efficiency is low, and precision is insufficient are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent scheduling of cascade reservoirs, and in particular to an intelligent scheduling method for cascade reservoirs. Background Art

[0002] In traditional cascade reservoir scheduling, scheduling decisions mainly rely on the experience judgment of schedulers and historical data. This method has the following main problems: Manual decision-making requires a large amount of time to analyze data and formulate scheduling plans, which may lead to slow response in case of emergencies and affect the scheduling effect. Due to the uncertainty of human factors, scheduling decisions may deviate, resulting in low accuracy and reliability of reservoir scheduling. Traditional methods are difficult to respond to the dynamic changes in reservoir operation in real time, such as rainfall, evaporation, water demand, etc., resulting in the scheduling plan may not be able to adapt to the actual situation changes. Traditional scheduling methods fail to make full use of modern information technologies, such as big data analysis and artificial intelligence algorithms, to mine the potential value in data and improve the scientificity and accuracy of scheduling. In the face of extreme climate and emergencies, traditional scheduling methods often lack effective coping strategies and are difficult to achieve fast and flexible scheduling of reservoir groups. With the rapid development of technologies such as artificial intelligence, big data, and cloud computing, these technologies provide new solutions for the intelligent scheduling of cascade reservoirs. By integrating these technologies, real-time monitoring, analysis, and prediction of reservoir operation data can be achieved, thereby improving the automation and intelligence level of scheduling. For example, machine learning algorithms can be used to predict key parameters such as reservoir inflow, outflow, and water level; genetic algorithms can be used to optimize scheduling rules and achieve multi-objective optimization; physical cause analysis can help understand the physical process of reservoir operation and improve the accuracy of the model; data mining techniques can discover potential laws in data and provide support for scheduling decisions.

[0003] Therefore, how to reduce manual intervention through intelligent methods, achieve fast response and automated scheduling, how to use advanced algorithms and models to improve the scientificity and accuracy of scheduling decisions, how to dynamically adjust the scheduling plan according to real-time data and prediction results to adapt to the dynamic changes of reservoir operation, how to mine the potential value in data through big data analysis and artificial intelligence technologies to improve the intelligence level of scheduling, how to quickly adjust the scheduling strategy in the face of extreme climate and emergencies to ensure the safety and benefits of reservoir groups, how to optimize the reservoir scheduling plan while meeting multiple objectives such as power generation, flood control, and irrigation to maximize economic and social benefits, and how to ensure that the scheduling system can still operate stably and provide reasonable scheduling plans in case of data loss or model prediction errors has become an urgent need. Summary of the Invention

[0004] The object of the present invention is to provide an intelligent scheduling method for cascade reservoirs to solve the technical problems in traditional cascade reservoir scheduling, such as relying on manual experience judgment, low decision-making efficiency and insufficient accuracy.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] An intelligent scheduling method for cascade reservoirs, comprising the following steps:

[0007] Step S1: Establish a comprehensive database, and periodically capture the scheduling data of each meteorological station, hydrological center, power plant and power grid for model call and calculation;

[0008] Step S2: Establish an intelligent simulation model, and based on the data in the database, conduct medium- and long-term runoff rolling forecasting and flood forecasting for the leading reservoir with seasonal regulation or above as the forecasting object, and conduct medium-term and flood forecasting for all power stations;

[0009] Step S3: The operator selects the scheduling object and the calculation time span according to the actual situation, adjusts the input and constraint conditions, and selects the appropriate optimization objective according to different scenario requirements;

[0010] Step S4: The intelligent simulation model adjusts the output and discharge of each subsequent period in real time according to the input information to achieve the set goal, outputs the flood control optimization scheduling plan, medium- and long-term scheduling plan and short-term scheduling plan, and supports manual intervention.

[0011] Further, the data in step S1 includes meteorological data, hydrological data, power plant operation data, unit operation data, power grid issued plans, and power grid load data.

[0012] Further, the input and constraint conditions in step S3 include system load constraints, inflow process, discharge limits, output range, water level range, maintenance process, shutdown, unit vibration range, and special scheduling instructions.

[0013] Furthermore, the flood control optimized operation plan includes a conventional operation method and an optimized operation mode; the conventional operation method is based on operation rules, utilizes the runoff regulation theory and the water energy calculation method, and implements flood control operation by means of the flood resistance capacity diagram of the reservoir and the empirical charts of the flood control operation diagram. Given the inflow flood process within the operation period, the initial water level of the reservoir, and the upper and lower limits of the water level during the flood regulation process, under the consideration of various constraints, the flood operation process of each reservoir is determined to make the water level of each time period of the reservoir meet the given upper and lower limit requirements as much as possible. Given the inflow flood process within the operation period, the initial water level of the reservoir, and the upper and lower limits of the outflow discharge during the flood regulation process, under the consideration of various constraints, the flood operation process of each reservoir is determined to make the outflow discharge of each time period of the reservoir meet the given upper and lower limit requirements as much as possible; in the flood control optimized operation, first, a corresponding model is established according to the highest water level, the maximum discharge, and the reservoir control water level at the end of the reservoir; at the same time, a fixed gate mode is added in the flood control operation, that is, after readjusting the constraints, the gate opening method and the opening degree remain unchanged, and an optimization model for gate adjustment is established based on this.

[0014] Furthermore, the model objective function of the highest water level of the reservoir is:

[0015]

[0016] In the formula: q(t) is the outflow process after being regulated by the reservoir; Q 区 (t) is the downstream interval flood process, the lowest highest water level model, and the end water level control model, which is obtained from the gate fixed model set in the fixed gate mode.

[0017] The model objective function of the maximum discharge is:

[0018] min Z max

[0019] In the formula: Z max is the highest water level during the flood regulation process;

[0020] The model objective function of the reservoir control water level at the end is:

[0021]

[0022] In the formula: Z n is the end water level of the flood regulation; Z end is the given end control water level of the flood regulation.

[0023] Furthermore, the objective function of the fixed gate mode model is:

[0024] min{max[Δz(θ1),Δz(θ2),······,Δz(θ n )]}(θ∈Ω)

[0025] where: θ is the feasible operation strategy; Ω is the set of feasible operation strategies; Δz(θ n ) is the water level change range of the reservoir under the feasible operation strategy θ.

[0026] Furthermore, the medium- and long-term scheduling plan is obtained through the maximum power generation model, the maximum power generation benefit model, the maximum cascade energy storage model, or the maximum comprehensive benefit model.

[0027] Furthermore, the objective function of the maximum power generation model is as follows:

[0028]

[0029] where: E is the objective function of maximum power generation; Δ t represents the number of hours in the t-th period; t and T represent the scheduling period number and its total number; m and M represent the reservoir number and the total number; p m,t represents the average output of the m-th reservoir in the t-th period;

[0030] The objective function of the maximum power generation benefit model is as follows:

[0031]

[0032] where: is the electricity price at time t of the m-th power station, and the others are the same as the maximum power generation model;

[0033] The objective function of the maximum cascade energy storage model is as follows:

[0034]

[0035] where: M is the total number of power stations participating in the calculation in the hydropower station group (1 ≤ m ≤ M); t is the period label, T is the total number of calculation periods, and 1 ≤ t ≤ T within the control period; El m represents the lag time electricity generated by the upstream power station of the m-th power station at the m-th power station.

[0036] Furthermore, the objective function of the maximum comprehensive benefit model is as follows:

[0037]

[0038] where: when the g-th objective is to find the minimum value when finding the maximum value is the maximum and minimum value of the g-th comprehensive benefit objective in the candidate solution, F i,g is the function value of the g-th comprehensive benefit objective of the i-th candidate solution, w g is the weight of each objective.

[0039] Further, the short-term scheduling scheme is obtained through the "electricity-determined water" model, the "water-determined electricity" model, or the peak shaving electricity quantity maximization model.

[0040] Due to the adoption of the above technical solution, the present invention has the following beneficial effects:

[0041] 1. The present invention periodically captures data and aggregates it into a comprehensive database. Through the intelligent simulation model, according to the input information, it adjusts the output and reservoir release in subsequent periods in real time to achieve the set goals, outputs the flood control optimal scheduling scheme, medium- and long-term scheduling scheme, and short-term scheduling scheme, and supports manual intervention; realizes rapid response and automated scheduling, uses advanced algorithms and models to improve the scientificity and accuracy of scheduling decisions, can quickly adjust the scheduling strategy, ensure the safety and benefits of the reservoir group, and how to optimize the reservoir scheduling scheme while meeting multiple goals such as power generation, flood control, and irrigation, realize the maximization of economic and social benefits, and ensure that the scheduling system can still operate stably and provide a reasonable scheduling scheme in the case of data loss or model prediction errors. Solve the technical problems existing in traditional cascade reservoir scheduling, such as relying on manual experience judgment, with low decision-making efficiency and insufficient accuracy. Description of the Drawings

[0042] Figure 1 It is a flowchart of the present invention. Detailed Embodiments

[0043] The following further describes the specific implementation of the invention with reference to the drawings.

[0044] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "length", "width", "upper", "lower", "vertical", "horizontal", "top", "bottom", "inner", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0045] In the present invention, unless otherwise clearly defined and limited, the terms "installed", "connected", "connected", "fixed", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can be directly connected, or indirectly connected through an intermediate medium, and can be the internal connection of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0046] In the present invention, unless otherwise clearly specified and defined, the first feature being "above" or "below" the second feature may include the direct contact between the first and second features, or may include the case where the first and second features are not in direct contact but in contact through additional features therebetween. Moreover, the first feature being "above", "over" and "on top of" the second feature includes the first feature being directly above and obliquely above the second feature, or merely indicating that the horizontal height of the first feature is higher than that of the second feature. The first feature being "below", "under" and "beneath" the second feature includes the first feature being directly below and obliquely below the second feature, or merely indicating that the horizontal height of the first feature is lower than that of the second feature.

[0047] As Figure 1 shown, a method for intelligent scheduling of cascade reservoirs includes the following steps:

[0048] Step S1: Establish a comprehensive database, periodically capture the data of each meteorological station, hydrological center, power plant and power grid dispatching, so as to prepare for the model call and calculation; the data includes meteorological data, hydrological data, power plant operation data, unit operation data, power grid issued plans, and power grid load data.

[0049] Step S2: Establish an intelligent simulation model and, based on the data in the database, conduct medium- and long-term runoff rolling forecasting and flood forecasting for the leading reservoir with a seasonal regulation performance or above as the forecasting object, and conduct medium-term and flood forecasting for all power stations;

[0050] Step S3: The operator selects the dispatching object and the calculation time span according to the actual situation, adjusts the input and constraint conditions, and selects the appropriate optimization objective according to the requirements of different scenarios; the input and constraint conditions include system load constraints, inflow process, outflow limits, output range, water level range, maintenance process, shutdown, unit vibration range, and special dispatching instructions.

[0051] Step S4: The intelligent simulation model adjusts the output and discharge in subsequent periods in real time according to the input information to achieve the set goals, outputs the flood control optimal operation plan, medium- and long-term operation plan, and short-term operation plan, and supports manual intervention. Specifically, the flood control optimal operation plan includes a conventional operation method and an optimal operation mode; the conventional operation method is based on operation rules, uses the runoff regulation theory and water energy calculation methods, and implements flood control operation with the help of the flood resistance capacity diagram and empirical charts of the flood control operation diagram of the reservoir. Given the inflow flood process, the initial water level of the reservoir, and the upper and lower limits of the water level during the flood regulation process within the given operation period, and considering various constraints, determine the flood regulation process of each reservoir so that the water level of each reservoir meets the given upper and lower limit requirements as much as possible. Given the inflow flood process, the initial water level of the reservoir, and the upper and lower limits of the discharge during the flood regulation process within the given operation period, and considering various constraints, determine the flood regulation process of each reservoir so that the discharge of each reservoir meets the given upper and lower limit requirements as much as possible; in the flood control optimal operation, first establish a corresponding model based on the highest water level, the maximum discharge, and the reservoir control water level at the end of the reservoir; at the same time, add a fixed gate mode in the flood control operation, that is, after readjusting the constraints, the gate opening method and opening remain unchanged, and establish an optimal model for gate adjustment accordingly.

[0052] Given the inflow discharge process and the initial water level of the reservoir within the given operation period, and considering various constraints, determine the flood regulation process of each reservoir to minimize the peak discharge downstream of the reservoir to the greatest extent. This model mainly considers that on the premise of meeting the flood control safety of the dam (or reservoir area), that is, making full use of the flood control storage capacity safely, try to meet the downstream flood control requirements and minimize the maximum value of the sum of the reservoir discharge and the downstream section. Specifically, the objective function of the model for the highest water level of the reservoir is:

[0053]

[0054] In the formula: q(t) is the discharge process after reservoir regulation; Q 区 (t) is the downstream section flood process, the lowest highest water level model, the final water level control model, which is obtained from the gate fixed model set in the fixed gate mode;

[0055] Given the inflow discharge process and the initial water level of the reservoir within the given operation period, and considering various constraints, determine the flood regulation process of each reservoir so that the highest water level during the flood regulation process of the reservoir is the lowest. The main purpose of this model is to reduce the flood control risk of the dam and the inundation loss of the reservoir area while meeting the safe discharge of the downstream flood control object to ensure the flood control safety of the dam; the objective function of the model for the maximum discharge is:

[0056] min Z max

[0057] In the formula: Z maxThe highest water level during the flood regulation process

[0058] The objective function of the reservoir control water level model at the end is:

[0059]

[0060] In the formula: Z n is the water level at the end of flood regulation; Z end is the given control water level at the end of flood regulation.

[0061] In this embodiment, given the inflow flood process and the initial reservoir water level during the scheduling period, considering various constraint conditions, combining the magnitude of the flood flow and the flood control storage capacity of the reservoir itself, the optimal gate opening scheme for each reservoir is determined, so that as the inflow changes, the gate opening method remains unchanged and can automatically meet the flood control requirements, and the water level fluctuation of the reservoir is minimized. The essence of the above goal is: in regulating a flood, under the same operation strategy, the difference between the maximum and minimum water levels among the reservoir water levels at each time period is taken, and the minimum value is taken among these maximum values under different operation strategies; the objective function of the fixed gate mode model is:

[0062] min{max[Δz(θ1),Δz(θ2),······,Δz(θ n )]}(θ∈Ω)

[0063] In the formula: θ is the feasible operation strategy; Ω is the set of feasible operation strategies; Δz(θ n ) is the water level change range of the reservoir under the feasible operation strategy θ. The main constraint conditions are as follows: (1) flood control storage capacity constraint; (2) water balance condition; (3) control end water level limit; (4) power generation diversion flow limit; (5) downstream discharge constraint; (6) reservoir water level constraint; (7) reservoir safe discharge constraint; (8) reservoir discharge capacity constraint; (9) reservoir discharge stability constraint; (10) calculated interval downstream discharge constraint; (11) river water balance limit. The above calculations can be automatically selected, manually set or forced according to the actual selected target. The detailed conditions are as follows:

[0064] Flood control storage capacity constraint:

[0065] Water balance constraint: V t = V t-1 +(I t -q t )·Δt

[0066] Control end water level limit: Z end ≤Z 预期

[0067] Power generation diversion flow limit: Q min ≤Qt,发电 ≤Q max

[0068] Downflow constraint: q min ≤q t ≤q max (1.5.9)

[0069] q min is the minimum flow rate required for comprehensive utilization; q max is the maximum flood discharge capacity limit.

[0070] Reservoir storage capacity constraint: Z min ≤Z t ≤Z max

[0071] Reservoir discharge capacity constraint: q t ≤Q(Z(T))

[0072] Reservoir maximum downflow constraint: q t ≤min(q 安 ,α*I 峰值 )

[0073] q 安 is the safe downstream discharge; α is the peak discount coefficient;

[0074] Reservoir discharge stability constraint: |q t -q t-1 |≤ε1, where ε1 is the flow rate change limit;

[0075] Calculated interval downflow constraint: q t,查 is the calculated downflow at time period t;

[0076] River channel water volume balance limit: q n,t =C0(n)I n-1,t +C1(n)I n-1,t-1 +C2(n)q n,t-1

[0077] In this paper, the Muskingum method is used for river channel flood routing, and the above formula is the water volume balance constraint of the nth river channel. In the formula, C0(n), C1(n), and C2(n) are the Muskingum flow routing coefficients of the nth river reach respectively, I n-1,t , I n-1,t-1 is the inflow at time t and (t - 1) of the nth river reach, q n,t , q n,t-1 is the outflow at time t and (t - 1) of the nth river reach. When conducting continuous multi-reach flood routing, the outflow of the (n - 1)th river reach is the inflow of the nth river reach. According to the characteristics of the Muskingum method, C0(n)+C1(n)+C2(n)=1.0.

[0078] In this embodiment, the medium- and long-term scheduling plan is obtained through the maximum power generation model, the maximum power generation benefit model, the maximum cascade energy storage model or the maximum comprehensive benefit model.

[0079] Given the inflow process into the reservoir and the initial and final water levels of the reservoir within the scheduling period, and considering various constraints, determine the scheduling process of each hydropower station reservoir with long-term regulation capacity, so as to maximize the system power generation (including the short-term lag power E M , and this item is taken as 0 for medium- and long-term). This model is the most commonly used optimal scheduling model for hydropower station groups. Its purpose is to make full use of the regulation capacity of hydropower stations as much as possible, increase the average power generation head and reduce the abandoned water. Especially, make full use of the hydraulic connection of cascade power stations and the power connection between basins to maximize the utilization of water energy resources under the consideration of the system's electricity load. The maximum power generation model can be divided into three different situations according to the given final water level, the given system energy storage at the end of the period, and the given system energy storage at the end of all or part of the periods. The latter two situations are mainly applied to medium- and long-term scheduling. Specifically, the objective function of the maximum power generation model is as follows:

[0080]

[0081] In the formula: E is the objective function of maximum power generation; Δ t represents the number of hours in the t-th period; t and T represent the scheduling period number and its total number; m and M represent the reservoir number and the total number; p m,t represents the average output of the m-th reservoir in the t-th period;

[0082] The power generation benefit of the maximum power generation benefit model is the product of the power generation and the average electricity price. The maximum power generation benefit model is basically the same as the maximum power generation model. The only difference is the introduction of the electricity price factor. Set the initial and final water levels of each reservoir during the scheduling period, with the goal of maximizing the power generation benefit during the scheduling period. The constraint conditions are the same as those of the maximum power generation model, and algorithms such as discrete differential dynamic programming and progressive optimization can be used to solve it. The objective function is as follows:

[0083]

[0084] In the formula: is the electricity price at the t-th moment of the m-th power station, and the others are the same as the maximum power generation model;

[0085] Under the condition that the power generation requirements of each period in the hydropower system are certain, given the initial water levels of each reservoir and the runoff of each period at the beginning of the scheduling, and under the constraints of the water level, the outflow flow, and the power generation output limit of the hydropower station, as well as the total output process of the hydropower station group, etc., use methods such as Lagrangian relaxation combination and direct load distribution to solve, reasonably distribute the load among each hydropower station, try to reduce the power generation water use, raise the power generation head, and increase the system energy storage, creating conditions for the safe, stable and economic operation of the future hydropower system. The objective function is as follows:

[0086]

[0087] Where: M is the total number of power stations in the hydropower station group participating in the calculation (1 ≤ m ≤ M); t is the time period label, T is the total number of calculation time periods, and 1 ≤ t ≤ T during the control period; El m represents the delay power generated by the upstream power station of power station m at power station m, and is calculated by the following formula:

[0088]

[0089] represents the reservoir capacity of reservoir m at the beginning of the t-th time period; η m represents the average water consumption rate of power station m; WT(m) represents the water storage volume above the dead water level at the end of the reservoir operation of power station m and all its upstream power station reservoirs; U m represents the array of labels of the direct upstream power stations of power station m; K m is the number of direct upstream power stations of the reservoir of hydropower station m.

[0090] In this embodiment, the cascade hydropower station group has comprehensive utilization requirements, corresponding to different optimal dispatching objectives, and the power generation dispatching itself can also be targeted at multiple objectives. The comprehensive benefit maximization model takes into account multiple power generation and comprehensive utilization objectives such as maximum power generation benefit, maximum power generation, minimum water consumption rate of power stations, ecological flow target, navigation requirements, etc. Taking ecological flow, navigation and other requirements as constraint conditions, a multi-objective optimization problem with other g objective functions is transformed into a single-objective optimization problem with fuzzy optimal membership degree as the objective. The objective function of the comprehensive benefit maximization model is as follows:

[0091]

[0092] In the formula: when the g-th objective is to find the minimum value when finding the maximum value is the maximum and minimum value of the g-th comprehensive benefit objective in the candidate solutions, F i,g is the function value of the g-th comprehensive benefit objective of the i-th candidate solution, w g is the weight of each objective.

[0093] Constraint conditions: The constraint conditions of the medium- and long-term power generation dispatching model of the hydropower station group generally include the following types:

[0094] Water level constraint: The water level above the dam of the reservoir should meet the following requirements for each time period during the control period: In the formula: Z m,t , Z m,t are the water level of power station m at the t-th time period and its upper and lower limits, unit: m.

[0095] Power output constraint of power station: The power output of the power station in each period during the control period should satisfy: Where: p m,t , p m,t is the power output of power station m in the t-th period and its upper and lower limits, unit: MW.

[0096] Power generation flow constraint: The power generation flow in each period should satisfy: Where: q m,t , q m,t is the power generation flow of power station m in the t-th period and its upper and lower limits, unit: m 3 / s.

[0097] Water volume balance constraint: The water volume balance constraint should be satisfied in each period: V m,t+1 = V m,t +(Q m,t - q m,t - Qd m,t )Δt; Where: V m,t and V m,t+1 respectively represent the reservoir capacities corresponding to the initial and final water levels in the t-th period, Q m,t , Qd m,t represent the incoming water flow and the discharged water flow of power station m in the t-th period, unit: m 3 / s.

[0098] Discharge flow constraint: The discharge flow in each period should satisfy and there is Q' m,t = q m,t + Qd m,t ; Where: Q' m,t , Q ' m,t is the discharge flow of power station m in the t-th period and its upper and lower limits, unit: m 3 / s.

[0099] Cascade energy storage constraint: The energy storage of key nodes is constrained according to the requirements of power grid safety, peak shaving, and coordination between thermal and hydro power (e.g., cascade energy storage at the beginning of the flood season and at the end of the year): Where: Es m,t , Es m,t is the energy storage at the end of the t-th period of power station m and its control upper and lower limits, unit: 100 million kWh.

[0100] In this embodiment, the short-term scheduling plan is obtained through the "electricity-determining water" model, the "water-determining electricity" model, or the peak shaving power maximization model.

[0101] Specifically, the "electricity-determining water" model is the cascade energy storage maximization model and the minimum power generation water consumption model;

[0102] When the power generation requirements of each time period in the hydropower system are certain, given the initial reservoir levels and runoff of each time period at the beginning of the dispatching, under the constraints of the water levels, out - flow discharges, power generation capacity limits of the hydropower stations, and the total power output process of the hydropower station group, methods such as Lagrangian relaxation combination and direct load distribution are used to solve, rationally distribute the load among the hydropower stations, minimize the power generation water consumption as much as possible, raise the power generation head, increase the system energy storage, and create conditions for the safe, stable and economic operation of the hydropower system in the future; the objective function of the maximum cascade energy storage model is as follows:

[0103]

[0104] Where: M is the total number of power stations participating in the calculation in the hydropower station group (1 ≤ m ≤ M); t is the time - period label, T is the total number of calculation time - periods, and 1 ≤ t ≤ T during the control period; ES m represents the electric energy that can be generated by the water volume above the dead water level of the m - th power station and all its upstream power stations at the m - th power station, and is calculated by the following formula; El m represents the time - lag electric energy generated by the upstream power stations of the m - th power station at the m - th power station.

[0105] When the load assigned to the hydropower station by the power system at a certain moment is certain, the optimal load distribution among the operating units of the hydropower station should satisfy the criterion of the minimum total working flow. The mathematical model of the minimum water consumption model is as follows:

[0106]

[0107] Where, the power plant load is when optimizing the load distribution between the 1 - st to i - th units, the total working flow of the whole plant; Q i (N i ,h i ): the working flow when the load of the i - th unit is N i ; Boundary condition, that is, the working flow is zero before the starting stage.

[0108] The "electricity determined by water" model is the maximum power generation model or the maximum power generation benefit model;

[0109] Given the inflow - discharge process and the initial and final reservoir levels within the dispatching period, under the consideration of various constraints, determine the dispatching process of each hydropower station reservoir, so that the system power generation is maximized (including the short - term time - lag electric energy E M ). The maximum power generation model can be divided into three different situations according to the given final water level, the given system energy storage at the end of the period, and the given system energy storage at the end of all or some time periods. The given final water level is mainly applied to short - term power generation dispatching; the objective function of the maximum power generation model is as follows:

[0110]

[0111] Where: E is the objective function for maximizing power generation; Δ t represents the number of hours in the t-th time period; t and T represent the scheduling period number and its total number; m and M represent the reservoir number and the total number; p m,t represents the average output of the m-th reservoir in the t-th time period, and E m is the lagging electricity of the m-th power station.

[0112] The power generation benefit is the product of the power generation and the average electricity price. The maximum power generation benefit model is basically the same as the maximum power generation model. The only difference is the introduction of the electricity price factor. Set the initial and final water levels of each reservoir during the scheduling period, with the goal of maximizing the power generation benefit during the scheduling period. The constraint conditions are the same as those of the maximum power generation model, and algorithms such as discrete differential dynamic programming and progressive optimization can be used to solve it. The objective function of the maximum power generation benefit model is as follows:

[0113]

[0114] Where: r m,t is the electricity price at time t of the m-th power station, and the others are the same as the maximum power generation model

[0115] Given the inflow process and the initial and final water levels of the reservoir during the scheduling period, under the consideration of various constraint conditions, determine the scheduling process of each hydropower station reservoir to maximize the peak shaving output of the system. This model is mainly applied to short-term scheduling. Its purpose is to consider the system load demand, make full use of the flexible operation characteristics of hydropower units, try to undertake the peak shaving task, and reduce the peak shaving burden of thermal power; the objective function of the peak shaving electricity maximum model:

[0116]

[0117] Among them, σ represents the variance of the remaining load, and C t represents the system load in the t-th time period, represents the output of the m-th power station in the t-th time period.

[0118] The water balance constraint of the short-term scheduling model is:

[0119] V m,t+1 = V m,t + 3600(Q m,t - q m,t - Qd m,t )Δ t

[0120] Where, Q m,t , q m,t and Qd m,t are respectively the inflow, generation flow and spillage flow of the m-th power station in the t-th time period; V m,t+1 , Vm,t respectively represent the reservoir storage of the m-th power station at the end and the beginning of the t-th time period.

[0121] Power generation flow constraint: q m,t ≤q m,t ≤q m,t ;

[0122] Reservoir water level constraint:

[0123] Outflow constraint:

[0124] Power station output constraint: p m,t ≤p m,t ≤p m,t ;

[0125] Minimum starting output constraint: In the formula, is the specified minimum output of the m-th power station, that is, P m,t is either greater than or equal to 0.

[0126] Power station output ramp-up constraint: In the formula, is the maximum available ramp-up output of the m-th power station in adjacent time periods.

[0127] Output duration constraint: (P m,t-α -P m,t-α-1 )(P m,t -P m,t-1 )≥0, α = 1, 2, …, t m -1; In the formula, t m is the number of time periods that the m-th power station needs to maintain at the highest or lowest point during a round of output increase and decrease, t m > 1.

[0128] Unit vibration zone constraint: In the formula, Ps k m,t are respectively the upper limit and the lower limit of the restricted output of the k-th output vibration zone existing in the m-th power station, that is, there should be and

[0129] Hydropower system bandwidth constraint: In the formula: h t represents the upper and lower limits of the average output of the hydropower system.

[0130] Grid partition section constraint: In the formula, It is the sum of the outputs of all large hydropower stations under the i-th section during the t-th period. It is the total sum of the outputs of each sub-region under the i-th section during the t-th period. It is the sum of the outputs of small hydropower stations under the i-th section during the t-th period. The calculation method of the total output of the sub-region is similar to that of the upper-level i-th region. It can be seen that the calculation process of the total output of a certain region is a recursive process until the smallest-level sub-region.

[0131] The above description is a detailed description of the preferred and feasible embodiments of the present invention. However, the embodiments are not intended to limit the scope of the patent application of the present invention. Any equivalent changes or modifications made under the technical spirit disclosed by the present invention should fall within the scope of the patent covered by the present invention.

Claims

1. An intelligent scheduling method for cascade reservoirs, characterized in that: It includes the following steps: Step S1: Establish a comprehensive database, and periodically capture data from each meteorological station, hydrological center, power plant, and power grid dispatching for model call and calculation; Step S2: Establish an intelligent simulation model and, based on the data in the database, conduct medium- and long-term runoff rolling forecasts and flood forecasts for the leading reservoirs with seasonal regulation and above as the forecasting objects, and conduct medium-term and flood forecasts for all power stations; Step S3: The operator selects the dispatching object and the calculation time span according to the actual situation, adjusts the input and constraint conditions, and selects appropriate optimization objectives according to different scenario requirements; Step S4: The intelligent simulation model adjusts the output and discharge of each subsequent period in real time according to the input information to achieve the set objectives, outputs flood control optimal dispatching plans, medium- and long-term dispatching plans, and short-term dispatching plans, and supports manual intervention.

2. The intelligent scheduling method for cascade reservoirs according to claim 1, wherein: The data in Step S1 includes meteorological data, hydrological data, power plant operation data, unit operation data, power grid issued plans, and power grid load data.

3. The intelligent scheduling method for cascade reservoirs according to claim 1, wherein: The input and constraint conditions in Step S3 include system load constraints, inflow process, discharge limits, output range, water level range, maintenance process, shutdown, unit vibration range, and special dispatching instructions.

4. The intelligent scheduling method for a cascade reservoir according to claim 1, wherein: The flood control optimal dispatching plan includes a conventional dispatching method and an optimal dispatching method; the conventional dispatching method is based on dispatching rules, uses the runoff regulation theory and hydropower calculation methods, and implements flood control dispatching operations with the help of the flood resistance capacity diagram of the reservoir and empirical charts of flood control dispatching diagrams. Given the inflow flood process during the dispatching period, the initial water level of the reservoir, and the upper and lower limits of the water level during the flood regulation process, considering various constraint conditions, determine the flood dispatching process of each reservoir so that the water level of each reservoir meets the given upper and lower limit requirements as much as possible. Given the inflow flood process during the dispatching period, the initial water level of the reservoir, and the upper and lower limits of the discharge flow during the flood regulation process, considering various constraint conditions, determine the flood dispatching process of each reservoir so that the discharge flow of each reservoir meets the given upper and lower limit requirements as much as possible; in the flood control optimal dispatching, first establish a corresponding model based on the highest water level of the reservoir, the maximum discharge, and the reservoir control water level at the end; at the same time, add a fixed gate mode in the flood control dispatching, that is, after readjusting the constraint conditions, the gate opening method and opening degree remain unchanged, and establish an optimal model for gate adjustment accordingly.

5. The intelligent scheduling method for cascade reservoirs according to claim 4, characterized in that: The model objective function for the highest water level of the reservoir is: (No interval) (with intervals) Where: q(t) is the outflow process after reservoir regulation; Q 区 (t) is the downstream reach flood process, the highest water level minimum model, and the end water level control model, which are obtained from the gate fixed model set in the fixed gate mode; The model objective function for the maximum discharge is: min Z max Where: Z max is the highest water level during the flood regulation process; The model objective function for the reservoir control water level at the end is: Where: Z n is the flood regulation end water level; Z end is the given flood regulation end control water level.

6. The intelligent scheduling method for a cascade reservoir according to claim 4, wherein: The model objective function for the fixed gate mode model is: min{max[Δz(θ1),Δz(θ2),······,Δz(θ n )]} (θ∈Ω) where: θ is the feasible operation strategy; Ω is the set of feasible operation strategies; Δz(θ n ) is the water level change range of the reservoir under the feasible operation strategy θ.

7. The intelligent scheduling method for cascade reservoirs according to claim 1, characterized in that: The medium- and long-term dispatching plan is obtained through the maximum power generation model, the maximum power generation benefit model, the maximum cascade energy storage model, or the maximum comprehensive benefit model.

8. The intelligent scheduling method for cascade reservoirs according to claim 7, characterized in that: The objective function of the maximum power generation model is as follows: Where: E is the maximum power generation objective function; Δ t represents the number of hours in the t-th period; t and T represent the scheduling period number and its total number; m and M represent the reservoir number and the total number; p m, t represents the average output of the m-th reservoir in the t-th period; The objective function of the maximum power generation benefit model is as follows: Where: is the electricity price at time t of the mth power station, and the others are the same as the maximum power generation model; The objective function of the maximum cascade energy storage model is as follows: Where: M is the total number of power stations in the hydropower station group participating in the calculation (1 ≤ m ≤ M); t is the time period label, T is the total number of calculation time periods, and 1 ≤ t ≤ T during the control period; El m represents the time-lag power generated by the upstream power station of the m-th power station at the m-th power station.

9. The intelligent scheduling method for a cascade reservoir according to claim 7, characterized in that: The objective function of the maximum comprehensive benefit model is as follows: where: when the g-th objective is to minimize when maximizing is the maximum and minimum value of the g-th comprehensive benefit objective among the candidate solutions, F i,g is the function value of the g-th comprehensive benefit objective of the i-th candidate solution, w g is the weight of each objective.

10. The intelligent scheduling method for cascade reservoirs according to claim 1, characterized in that: The short-term dispatching plan is obtained through the "electricity-determined water" model, the "water-determined electricity" model, or the maximum peak shaving power model.

Images