Capacity optimization configuration method for seasonal water-fire-wind-light multi-energy complementary system

Abstract

The invention relates to a seasonal water-fire-wind-light multi-energy complementary system capacity optimization configuration method, which comprises the following steps of S1, selecting different optimization configuration schemes for flat, rich and dry seasons according to a seasonal generating capacity prediction result of a hydropower station in a multi-energy complementary system; s2, determining an expression of a target function, and establishing a multi-energy complementary system optimization model by taking the minimum investment cost limit value, the minimum net load fluctuation, the maximum new energy generating capacity and the minimum total carbon emission as sub-targets; s3, establishing constraint conditions; and S4, in different seasons, different optimal configuration schemes are adopted, and the optimal number of hydroelectric generating sets, the optimal number of photovoltaic arrays, the optimal number of wind generating sets and the optimal number of thermal power generating sets of the multi-energy capacity optimal configuration system are solved through a particle swarm optimization algorithm. According to the method, multiple sub-targets and constraint conditions are comprehensively considered, the benefits of various power supplies and power grid companies can be comprehensively considered, the energy cost is reduced, and the energy economy is improved.

Description

Technical Field

[0001] The present invention relates to the field of multi-energy complementary power generation, and in particular to a capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system. Background Art

[0002] The multi-energy complementary power generation system is based on the natural complementary characteristics of different renewable energy and conventional energy. It realizes a multi-energy hybrid power generation system and is one of the important contents of the construction of clean energy bases. The optimal configuration of the complementary system capacity is the key to achieving higher quality power output, reducing economic costs and maximizing comprehensive benefits.

[0003] At present, the mode of energy production and consumption is transforming towards green and low-carbon. New energy power generation methods such as photovoltaic power generation and wind power generation are characterized by high efficiency, low carbon and environmental protection, indirectness, and volatility. Thermal power and hydropower units are flexible in start and stop, and can quickly adjust the power fluctuations caused by the uncertainty of wind and solar power, thereby effectively improving the flexibility of the system and promoting the consumption of renewable energy.

[0004] At present, a large number of studies have been carried out on the optimal configuration of multi-energy complementary capacity, mainly involving optimization models in terms of reliability and economy. There are few studies that use the complementary system structure of water, fire, wind and light alone. Most of them only optimize the system configuration results from a single objective and single dimension, and rarely consider the impact of seasonal factors on various energy structures in multi-energy complementary systems. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system that comprehensively considers multiple objectives and constraints.

[0006] In order to solve the above problems, the present invention provides a method for optimizing the capacity of a seasonal water, fire, wind and solar multi-energy complementary system, comprising the following steps:

[0007] S1: According to the seasonal power generation forecast results of the hydropower station in the multi-energy complementary system, different optimization configuration schemes are selected for the normal, high and low water seasons;

[0008] S2: Determine the expression of the objective function, and then take the minimum investment cost limit, the minimum net load fluctuation, the maximum renewable energy power generation and the minimum total carbon emissions as sub-objectives, and perform Min-Max normalization on the sub-objective functions; then establish a multi-energy complementary system optimization model through linear weighted summation;

[0009] S3: Establish constraints:

[0010] The constraints refer to system single-machine installed capacity constraints, energy waste constraints, thermal power safety startup constraints, hydroelectric generator output constraints, photovoltaic generator output constraints, wind generator output constraints and thermal power output constraints;

[0011] S4: Under different seasonal conditions, different optimization configuration schemes are adopted, and the optimal number of hydropower generating units, photovoltaic arrays, wind power generating units and thermal power generating units of the multi-energy capacity optimization configuration system is solved by particle swarm optimization algorithm.

[0012] The seasonal power generation prediction result of the hydropower station in step S1 is obtained by the following method: first, the hydropower station data with complete information in the multi-energy complementary system is preprocessed; then, the seasonal power generation of the hydropower station is predicted using the LSTM model according to the characteristics of the hydropower station data; finally, the seasonal power generation capacity prediction result of the hydropower station is analyzed and divided into normal, flood and dry seasons.

[0013] The hydropower station data preprocessing refers to the process of collecting, sorting, standardizing and normalizing the hydropower station data including the water volume of power generation in previous years, performing correlation analysis and outlier processing.

[0014] The LSTM model consists of a forget gate f t , input gate i t and g t , cell state c t and output gate o t Composition, its expression is:

[0015] Forget Gate t Press the formula to obtain:

[0016] f t =σ(W f h t-1 +U f x t +b f )

[0017] Where: σ is the sigmoid activation function; W f and U f is the weight matrix of the forget gate; b f is the bias vector; hx -1 is the hidden state at the previous time t-1; x t is the value of the input sequence at the current time t;

[0018] Input Gate i t and g t According to the formula:

[0019] i t =σ(W i h t-1+U i x t +b i )

[0020] g t =tanh(W g h t-1 +U g x t +b g )

[0021] Where: W i , U i , W g and U g is the weight matrix of the input gate; b i and b g is the bias vector; tanh is the hyperbolic tangent activation function;

[0022] Cell state c t According to the formula:

[0023] c t =c t-1 ⊙f t +f t ⊙g t

[0024] Where: c t-1 is the cell state at the previous moment; ⊙ is the Hadamard product;

[0025] Output gate o t According to the formula:

[0026] o t =σ(W o h t-1 +U o x t +b o )

[0027] h t =o t ⊙tanh(c t )

[0028] Where: W o and U o is the weight matrix of the output gate; b o is the bias vector; h t is the hidden state at the current time t.

[0029] The optimization configuration scheme for normal, high and low water seasons in step S1 refers to:

[0030] During the flood season, hydropower is fully generated to maximize the output of hydropower. If wind power and thermal power are required to participate in regulation, they will be generated in the order of hydropower, wind power, solar power and thermal power.

[0031] During the dry season, the goal is to maximize the amount of wind and solar power generation. In order to avoid adverse effects on the downstream ecological environment during the power generation process of the hydropower station, the thermal power and hydropower downstream ecological flow are used to participate in the regulation as the base load of the power system. The load of the power system is generated in the order of wind and solar power, hydropower downstream ecological flow generation, thermal power and hydropower peak load regulation;

[0032] During the normal water period, the goal is to maximize wind and solar power generation, and power is generated in the order of wind, solar, hydropower and thermal power.

[0033] The multi-energy complementary system optimization model in step S2 is determined by the following method:

[0034] ①Determine the expression of the objective function:

[0035] min Z=λ1f1+λ2f2+λ3f3++λ4f4

[0036] Where: f1 is the investment cost limit C; f2 is the load power shortage rate D; f3 is the renewable energy power generation W; f4 is the total carbon emissions Q; where f1 = C, f2 = D, f3 = -W, f4 = Q; λ1, λ2, λ3 and λ4 are the weight coefficients of the investment cost limit, load power shortage rate, renewable energy total power generation and total carbon emissions, respectively, and λ1+λ2+λ3+λ4=1;

[0037] ② Determine the investment cost sub-target, which is expressed as:

[0038] min f1=C1+C2+C3;C1=C s N s S s +C pv N pv S pv +C w N w S w +C h N h S h ; C2 = γC1;

[0039]

[0040] Where: C1 is the system equipment investment cost; C2 is the system equipment operation and maintenance cost; C3 is the coal consumption and start-up and shutdown cost of the thermal power unit; C s , C pv , C w and C hare the unit capacity investment costs of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; N s 、N pv 、N w and N h are the number of hydropower units, photovoltaic arrays, wind power units and thermal power units respectively; S s , S pv , S w and S h are the total capacities of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; γ is the operation and maintenance cost coefficient; C mh is the coal consumption cost of thermal power units; C q is the start-up and shutdown cost of the thermal power unit; T is the unit output dispatch cycle; P h (t) is the output power of a single thermal power unit at time t; a, b, c are the consumption characteristic coefficients of the thermal power unit; u t A is the start and stop status of the thermal power unit; h is the start-up and shutdown cost of thermal power units;

[0041] ③ Determine the minimum net load fluctuation sub-goal, which is expressed as:

[0042]

[0043] Where: Pgh is the net load value at time t; Pgh,av is the average net load value within a dispatching cycle;

[0044] ④ Based on seasonal factors, consider the maximum sub-goal of renewable energy power generation in three situations: normal, high and low water seasons:

[0045] During the flood season, hydropower is fully generated, and its sub-objective expression is:

[0046]

[0047] During the dry season, the wind and solar power generation is the largest, and its sub-objective expression is:

[0048]

[0049] During the flat water period, the wind and solar power generation is the largest, and its sub-objective expression is:

[0050]

[0051] Where: T is the unit output dispatch cycle; P s (t), P pv (t) and P w (t) are the output power of a single hydropower unit, a unit photovoltaic array, and a single wind turbine unit at time t;

[0052] ⑤ Determine the total carbon emission quantum target, the expression is:

[0053] max f4=λE C

[0054] Where: λ is the carbon emission factor; E C is the total thermal power generation;

[0055] ⑥ Perform Min-Max normalization on the four sub-targets f1, f2, f3, and f4 respectively:

[0056]

[0057] ⑦ Perform linear weighted summation on the sub-goals and establish a multi-energy complementary system optimization model:

[0058] min Z′=λ1f1′+λ2f′2+λ3f′3+λ4f′4

[0059] Where: f′1, f′2, f′3 and f′4 are the normalized values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1max 、f 2max 、f 3max 、f 4max are the maximum values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1min 、f 2min 、f 3min 、f 4min They are respectively the minimum values ​​of investment cost limit, load power shortage rate, new energy power generation and total carbon emissions.

[0060] The constraint conditions in step S3 are established as follows:

[0061] (1) System single machine installed capacity constraints:

[0062]

[0063] Where: Z s , Z h , Z w , Z pv are the installed capacity of hydropower, thermal power, wind power and photovoltaic units respectively; Z smin , Z hmin , Z pvmin , Z wmin are the lower limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power generation and wind power generation in the target year; Z smax , Z hmax , Z pvmax , Z wmaxThey are the upper limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power and wind power in the target year;

[0064] (2) Energy waste rate constraint:

[0065]

[0066] Where: D WE is the energy waste rate; D WEmax is the maximum energy waste rate; P WE (t) is the energy wasted power at time t; E is the total load power consumption within T time;

[0067] (3) Thermal power safety startup restrictions:

[0068] H≥C hmin R hmin T max

[0069] Where: C hmin is the thermal power safety startup capacity obtained according to the regional power grid safety and stability constraints; R hmin is the minimum output rate of thermal power; T max is the number of hours in a year, 8760 hours in a non-leap year and 8784 hours in a leap year;

[0070] (4) Output constraints of hydroelectric generators, photovoltaic generators, wind generators and thermal power generation:

[0071]

[0072] Where: P smin is the minimum output power of the unit hydropower unit; P smax , P pvmax and P wmax is the maximum output power of hydropower, photovoltaic and wind turbine units; P h is the output power of the thermal power unit; H min , H max It is the minimum output coefficient and maximum output coefficient of thermal power units.

[0073] Compared with the prior art, the present invention has the following advantages:

[0074] 1. Based on the consideration of seasonal factors, the present invention adopts a water-fire-wind-solar multi-energy complementary power generation system based on hydropower clean energy, fully considers the power generation characteristics of hydropower in normal, abundant and dry seasons, combines the seasonal characteristics of thermal power, photovoltaic power generation and wind power, comprehensively considers multiple objectives and constraints, and constructs a seasonal multi-energy complementary system capacity optimization configuration system.

[0075] 2. The present invention comprehensively considers multiple sub-goals and constraints, and uses the particle swarm optimization algorithm to reasonably allocate and utilize energy according to the actual situation and needs of energy, avoids the impact of large-scale introduction of new energy on the benefits of traditional energy and power grid companies, and takes into account the benefits of various power sources and power grid companies, providing a basis for the coordinated development of new energy and conventional energy, thereby reducing energy costs and improving the economy of energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0077] Figure 1 This is a correlation test diagram of the power generation of a hydropower station in Gannan Prefecture over the years in an embodiment of the present invention. DETAILED DESCRIPTION

[0078] A capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system comprises the following steps:

[0079] S1: According to the seasonal power generation forecast results of the hydropower station in the multi-energy complementary system, different optimization configuration schemes are selected for the normal, abundant and dry seasons.

[0080] Among them: the seasonal power generation prediction results of the hydropower station are obtained according to the following method: first, the hydropower station data with complete information in the multi-energy complementary system are preprocessed; then the seasonal power generation of the hydropower station is predicted using the LSTM model according to the characteristics of the hydropower station data; finally, the seasonal power generation capacity prediction results of the hydropower station are analyzed and divided into normal, flood and dry seasons.

[0081] Hydropower station data preprocessing refers to the process of collecting, organizing, standardizing and normalizing hydropower station data including the amount of water generated over the years, performing correlation analysis, and handling outliers.

[0082] The LSTM model consists of a forget gate f t , input gate i t and g t , cell state c t and output gate o t Composition, its expression is:

[0083] Forget Gate t According to the formula:

[0084] f t =σ(W f h t-1 +U f x t +b f )

[0085] Where: σ is the sigmoid activation function; W f and Uf is the weight matrix of the forget gate; b f is the bias vector; h t-1 is the hidden state at the previous time t-1; x t is the value of the input sequence at the current time t;

[0086] Input Gate i t and g t According to the formula:

[0087] i t =σ(W i h t-1 +U i x t +b i )

[0088] g t =tanh(W g h t-1 +U g x t +b g )

[0089] Where: W i , U i , W g and U g is the weight matrix of the input gate; b i and b g is the bias vector; tanh is the hyperbolic tangent activation function;

[0090] Cell state c t According to the formula:

[0091] C t =C t-1 ⊙f t +f t ⊙g t

[0092] Where: c t-1 is the cell state at the previous moment; ⊙ is the Hadamard product;

[0093] Output gate o t According to the formula:

[0094] o t =σ(W o h t-1 +U o x t +b o )

[0095] h t =o t ⊙tanh(c t )

[0096] Where: W o and U o is the weight matrix of the output gate; b o is the bias vector; h t is the hidden state at the current time t.

[0097] The optimal allocation plan for normal, high and low water seasons refers to:

[0098] During the flood season, hydropower is fully generated with the goal of maximizing hydropower output. If wind power and thermal power are needed to participate in regulation, they will be generated in the order of hydropower, wind power, solar power and thermal power.

[0099] During the dry season, the goal is to maximize wind and solar power generation. In order to avoid adverse effects on the downstream ecological environment during the power generation process of the hydropower station, the thermal power and hydropower downstream ecological flow are used to participate in the regulation as the base load of the power system. The load of the power system is output in the order of wind and solar, hydropower downstream ecological flow power generation, thermal power and hydropower peak regulation.

[0100] During the normal water period, the goal is to maximize wind and solar power generation, and power is generated in the order of wind, solar, hydropower and thermal power.

[0101] S2: Determine the expression of the objective function, and then take the minimum investment cost limit, the minimum net load fluctuation, the maximum new energy power generation and the minimum total carbon emissions as sub-objectives, and perform Min-Max normalization on the sub-objective functions; then establish the multi-energy complementary system optimization model through linear weighted summation. The specific method is as follows:

[0102] ①Determine the expression of the objective function:

[0103] minZ=λ1f1+λ2f2+λ3f3++λ4f4

[0104] Where: f1 is the investment cost limit C; f2 is the load power shortage rate D; f3 is the renewable energy power generation W; f4 is the total carbon emissions Q; where f1 = C, f2 = D, f3 = -W, f4 = Q; λ1, λ2, λ3 and λ4 are the weight coefficients of the investment cost limit, load power shortage rate, renewable energy total power generation and total carbon emissions, respectively, and λ1+λ2+λ3+λ4=1;

[0105] ② Determine the investment cost sub-target, which is expressed as:

[0106] minf1=C1+C2+C3;C1=C s N s S s +C pv N pv S pv +Cw N w S w +C h N h S h ; C2 = γC1;

[0107] Where: C1 is the system equipment investment cost; C2 is the system equipment operation and maintenance cost; C3 is the coal consumption and start-up and shutdown cost of the thermal power unit; C s , C pv , C w and C h are the unit capacity investment costs of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; N s 、N pv 、N w and N h are the number of hydropower units, photovoltaic arrays, wind power units and thermal power units respectively; S s , S pv , S w and Sx are the total capacities of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; γ is the operation and maintenance cost coefficient; C mh is the coal consumption cost of thermal power units; C q is the start-up and shutdown cost of the thermal power unit; T is the unit output dispatch cycle; P h (t) is the output power of a single thermal power unit at time t; a, b, c are the consumption characteristic coefficients of the thermal power unit; u t A is the start and stop status of the thermal power unit; h is the start-up and shutdown cost of thermal power units;

[0108] ③ After the combined output of water, wind and solar power is subtracted from the entire system, the load borne by the thermal power unit is called the net load. To prevent the thermal power unit from frequently adjusting its output, the net load volatility borne by the thermal power unit should be reduced as much as possible. Therefore, the net load fluctuation minimum sub-goal is determined, and its expression is:

[0109]

[0110] Where: Pgh is the net load value at time t; Pgh,av is the average net load value within a dispatching cycle;

[0111] ④ Based on seasonal factors, consider the maximum sub-goal of renewable energy power generation in three situations: normal, high and low water seasons:

[0112] During the flood season, hydropower is fully generated, and its sub-objective expression is:

[0113]

[0114] During the dry season, the wind and solar power generation is the largest, and its sub-objective expression is:

[0115]

[0116] During the flat water period, the wind and solar power generation is the largest, and its sub-objective expression is:

[0117]

[0118] Where: T is the unit output dispatch cycle; P s (t), P pv (t) and P w (t) are the output power of a single hydropower unit, a unit photovoltaic array, and a single wind turbine unit at time t;

[0119] ⑤ Determine the total carbon emission quantum target, the expression is:

[0120] maxf4=λE C

[0121] Where: λ is the carbon emission factor; E C is the total thermal power generation;

[0122] ⑥ Perform Min-Max normalization on the four sub-targets f1, f2, f3, and f4 respectively:

[0123]

[0124] ⑦ Perform linear weighted summation on the sub-goals and establish a multi-energy complementary system optimization model:

[0125] minZ′=λ1f′1+λ2f′2+λ3f′3+λ4f′4

[0126] Where: f′1, f′2, f′3 and f′4 are the normalized values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1max 、f 2max 、f 3max 、f 4max are the maximum values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1min 、f 2min 、f 3min 、f 4min They are respectively the minimum values ​​of investment cost limit, load power shortage rate, new energy power generation and total carbon emissions.

[0127] S3: To avoid the maximum encroachment of photovoltaic and wind power generation on the living space of hydropower, thermal power and power grid companies, and to ensure local electricity consumption and external power transmission needs, as well as the safety and stability constraints of the power grid, we will coordinate and take into account the coordinated development and scientific allocation of various power sources, limit the planned installed capacity of various power sources, and establish constraints.

[0128] The constraints refer to the system single-machine installed capacity constraints, energy waste constraints, thermal power safety startup constraints, hydropower unit output constraints, photovoltaic unit output constraints, wind power unit output constraints and thermal power output constraints.

[0129] (1) System single machine installed capacity constraints

[0130]

[0131] Where: Z s , Z h , Z w , Z pv are the installed capacity of hydropower, thermal power, wind power and photovoltaic units respectively; Z smin , Z hmin , Z pvmin , Z wmin are the lower limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power and wind power in the target year; Z smax , Z hmax , Z pvmax , Z wmax They are the upper limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power and wind power in the target year;

[0132] (2) Energy waste rate constraint:

[0133]

[0134] Where: D WE is the energy waste rate; D WE max is the maximum energy waste rate; P WE (t) is the energy wasted power at time t; E is the total load power consumption within T time;

[0135] ⑶ In order to ensure the safe and stable operation of the power system, the thermal power startup capacity should be greater than or equal to the thermal power safety startup capacity. Converted into thermal power generation, the thermal power safety startup constraints are:

[0136] H≥C hmin R hmin T max

[0137] Where: C h minis the thermal power safety startup capacity obtained according to the regional power grid safety and stability constraints; R h min is the minimum output rate of thermal power; T max is the number of hours in a year, 8760 hours in a non-leap year and 8784 hours in a leap year;

[0138] (4) Output constraints of hydroelectric generators, photovoltaic generators, wind generators and thermal power generation:

[0139]

[0140] Where: P s min is the minimum output power of the unit hydropower unit; P s max , P pv max and P w max is the maximum output power of hydropower, photovoltaic and wind turbine units; P h is the output power of the thermal power unit; H min , H max It is the minimum output coefficient and maximum output coefficient of thermal power units.

[0141] S4: Solve the linear programming problem composed of the objective function and the constraints. In different seasonal conditions, different optimization configuration schemes are adopted, and the optimal number of hydropower generating units, photovoltaic arrays, wind power generating units and thermal power generating units of the multi-energy capacity optimization configuration system is solved by the particle swarm optimization algorithm.

[0142] In order to simplify the objective function and the solution of each sub-objective, the particle swarm optimization algorithm is introduced in the calculation process. The particle swarm algorithm abstracts the target solution of the problem to be optimized as the predation target of the bird flock. The basic principle of the particle swarm comes from the imitation of the foraging behavior of the bird flock. A specific solution of the optimization problem is simulated as a bird that can fly freely in a high-dimensional infinite space, that is, a basic particle. The position of the particle in the search space is the current solution of the current particle to the optimization problem. For each solution, the particle swarm algorithm uses fitness to evaluate its quality. The optimal solution of all particles at all times and the optimal solution of each particle at all times in the search space The corresponding two positions determine the optimization speed of each particle, and the optimal solution is found through continuous iteration of particle position and speed.

[0143] Taking seasonal factors into consideration, according to the seasonal characteristics of normal, high and low water seasons, different operation strategies are adopted, with the minimum investment cost limit, minimum net load fluctuation, maximum renewable energy power generation and minimum carbon emissions as the objective function. The specific steps of the particle swarm algorithm in the application of seasonal water, fire, light and wind multi-energy complementary system capacity optimization configuration are as follows:

[0144] (1) Initialize the particle swarm. According to the established capacity optimization configuration model of seasonal water, fire, light and wind multi-energy complementary system, determine the parameters of the particle swarm algorithm (such as population size, maximum evolutionary generations, inertia weight).

[0145] (2) The spatial dimension of the particle swarm is determined by the number of independent variables, and the position and velocity of the particle swarm are randomly initialized according to the power installed capacity constraint, energy waste constraint, thermal power safety startup constraint, hydropower unit output constraint, photovoltaic unit output constraint, wind power unit output constraint and thermal power output constraint.

[0146] (3) The fitness value of each particle is calculated by taking the minimum investment cost limit, the minimum net load fluctuation, the maximum new energy power generation and the minimum carbon emission as the fitness value of each particle.

[0147] (4) Compare the fitness value of each particle with the individual extreme value. If it is better, update the individual extreme value to the fitness value of the particle. Otherwise, determine whether to accept the deteriorated solution.

[0148] (5) Compare the fitness value of each particle with the global extreme value. If the fitness value is greater than the global extreme value, take the fitness value as the new global extreme value.

[0149] (6) Update the position and flight speed of each particle.

[0150] (7) Check whether the conditions for iterative termination have been met. If the current number of iterations reaches the maximum evolutionary generation or meets the error requirement, the iteration is terminated and the optimal number of hydropower units, wind turbines, photovoltaic arrays, and thermal power units is output. If not, go to step (3) to continue the iterative search.

[0151] (8) The iteration is terminated, and the position of the global optimal value point obtained is the optimal value of the seasonal water, fire, light and wind multi-energy complementary system capacity optimization configuration system.

[0152] Example

[0153] Taking a power station in Gannan Prefecture as an example, the method for predicting seasonal power generation of a hydropower station of the present invention is described in detail.

[0154] Gannan Prefecture is located in the upper reaches of the Yangtze River and the Yellow River. There are more than 120 main and tributary rivers in the territory, including the Yellow River, Tao River, Daxia River, Bailong River, etc. The theoretical reserves of hydropower resources are 3.61 million kilowatts, accounting for 21% of the province's hydropower resources, and the exploitable amount is 2.16 million kilowatts, accounting for 22% of the province's exploitable amount. The power generation of a hydropower station in Gannan Prefecture over the years is shown in Table 1.

[0155] Table 1 Annual power generation of a hydropower station in Gannan Prefecture Unit: 10,000 kWh

[0156]

[0157]

[0158] A correlation test was conducted on the power generation of a hydropower station in Gannan Prefecture over the years. Figure 1 shown.

[0159] Data normalization is one of the important steps in data preprocessing. Its purpose is to make the objective function converge quickly. The normalization operation is because the original dimension of the data is not suitable for the model input. The present invention normalizes the input and output data of each sample in advance. The formula is:

[0160]

[0161] Where: x i is the original data; x max is the maximum value; x min is the minimum value; i is the normalized value, and y i ∈[0,1].

[0162] Depend on Figure 1 It can be seen that the power generation of the hydropower station has good correlation, and the LSTM model can be used to predict the power generation. The specific process is as follows:

[0163] Step 1: Input the normalized historical power generation of the hydropower station into the LSTM model. There are 120 data in total for this hydropower station, 84 of which are set as training sets and 36 as test sets.

[0164] Step 2: Pass the forget gate f t The sigmoid function plays a major role in determining the information that the unit needs to discard, mapping the information value to the interval [0,1]. When the value is 0, it means that everything is forgotten, and when the value is 1, it means that everything is retained.

[0165] Step 3: Pass the input gate i t and g t Determine what information to retain from the input information, and then use the tanh activation function to obtain new candidate information and store it in the cell state c t middle.

[0166] Step 4: Set the cell state c t Update, through the forget gate f t Forget some information and pass the input gate i t and g t Add candidate unit information to obtain new cell units.

[0167] Step 5: Input the information into the new cell unit and determine the characteristics of the output unit by running a sigmoid function. Use the tanh activation function to map the cell state value to [-1,1], then multiply it by the judgment condition obtained by the sigmoid function and pass the information through the output gate o t Output.

[0168] The prediction results are shown in Tables 2 and 3:

[0169] Table 2 Forecast of power generation of hydropower stations in 2019 Unit: 10,000 kWh

[0170]

[0171] Table 3 Forecast of power generation of hydropower stations in 2020 Unit: 10,000 kWh

[0172]

[0173] Table 4 Forecast of power generation of hydropower stations in 2021 Unit: 10,000 kWh

[0174]

[0175] It can be seen from Tables 2 to 4 that the prediction results of the LSTM model can meet the accuracy requirements and can be applied in practice, enabling power grid companies to formulate multi-energy complementary system scheduling plans in advance based on the hydropower generation prediction results and formulate more accurate and reasonable electricity consumption plans.

Claims

1. A capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system, comprising the following steps: S1: According to the seasonal power generation forecast results of the hydropower station in the multi-energy complementary system, different optimization configuration schemes are selected for the normal, high and low water seasons; S2: Determine the expression of the objective function, and then take the minimum investment cost limit, the minimum net load fluctuation, the maximum renewable energy power generation and the minimum total carbon emissions as sub-objectives, and perform Min-Max normalization on the sub-objective functions; then establish a multi-energy complementary system optimization model through linear weighted summation; S3: Establish constraints: The constraints refer to system single-machine installed capacity constraints, energy waste constraints, thermal power safety startup constraints, hydroelectric generator output constraints, photovoltaic generator output constraints, wind generator output constraints and thermal power output constraints; S4: Under different seasonal conditions, different optimization configuration schemes are adopted, and the optimal number of hydropower generating units, photovoltaic arrays, wind power generating units and thermal power generating units of the multi-energy capacity optimization configuration system is solved by particle swarm optimization algorithm.

2. A capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system as claimed in claim 1, characterized in that: The seasonal power generation prediction result of the hydropower station in step S1 is obtained by the following method: first, the hydropower station data with complete information in the multi-energy complementary system is preprocessed; then, the seasonal power generation of the hydropower station is predicted using the LSTM model according to the characteristics of the hydropower station data; finally, the seasonal power generation capacity prediction result of the hydropower station is analyzed and divided into normal, flood and dry seasons.

3. A capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system as claimed in claim 2, characterized in that: The hydropower station data preprocessing refers to the process of collecting, sorting, standardizing and normalizing the hydropower station data including the water volume of power generation in previous years, performing correlation analysis and outlier processing.

4. A capacity optimization configuration method for a seasonal water, fire, wind and solar multi-energy complementary system as claimed in claim 2, characterized in that: The LSTM model consists of a forget gate f t , input gate i t and g t , cell state c t and output gate o t Composition, its expression is: Forget Gate t Press the formula to obtain: f t =σ(W f h t-1 +U f x t +b f ) Where: σ is the sigmoid activation function; W f and U f is the weight matrix of the forget gate; b f is the bias vector; h t-1 is the hidden state at the previous time t-1; x t is the value of the input sequence at the current time t; Input Gate i t and g t Press the formula to obtain: i t =σ(W i h t-1 +U i x t +b i ) g t =tanh(W g h t-1 +U g x t +b g ) Where: W i , U i , W g and U g is the weight matrix of the input gate; b i and b g is the bias vector; tanh is the hyperbolic tangent activation function; Cell state c t Press the formula to obtain: c t =c t-1 ⊙f t +f t ⊙g t Where: c t-1 is the cell state at the previous moment; ⊙ is the Hadamard product; Output gate o t Press the formula to obtain: the t =σ(W o h t-1 +U o x t +b o ) h t =o t ⊙tanh(c t ) Where: W o and U o is the weight matrix of the output gate; b o is the bias vector; h t is the hidden state at the current time t.

5. The capacity optimization configuration method of a seasonal water, fire, wind and solar multi-energy complementary system as claimed in claim 1, characterized in that: The optimization configuration scheme for normal, high and low water seasons in step S1 refers to: During the flood season, hydropower is fully generated to maximize the output of hydropower. If wind power and thermal power are required to participate in regulation, they will be generated in the order of hydropower, wind power, solar power and thermal power. During the dry season, the goal is to maximize the amount of wind and solar power generation. In order to avoid adverse effects on the downstream ecological environment during the power generation process of the hydropower station, the thermal power and hydropower downstream ecological flow are used to participate in the regulation as the base load of the power system. The load of the power system is generated in the order of wind and solar power, hydropower downstream ecological flow generation, thermal power and hydropower peak load regulation; During the normal water period, the goal is to maximize wind and solar power generation, and power is generated in the order of wind, solar, hydropower and thermal power.

6. The capacity optimization configuration method of a seasonal water, fire, wind and solar multi-energy complementary system according to claim 1, characterized in that: The multi-energy complementary system optimization model in step S2 is determined by the following method: ①Determine the expression of the objective function: min Z=λ1f1+λ2f2+λ3f3++λ4f4 Where: f1 is the investment cost limit C; f2 is the load power shortage rate D; f3 is the renewable energy power generation W; f4 is the total carbon emissions Q; where f1 = C, f2 = D, f3 = -W, f4 = Q; λ1, λ2, λ3 and λ4 are the weight coefficients of the investment cost limit, load power shortage rate, renewable energy total power generation and total carbon emissions, respectively, and λ1+λ2+λ3+λ4=1; ② Determine the investment cost sub-target, which is expressed as: min f1=C1+C2+C3;C1=C s N s S s +C pv N pv S pv +C w N w S w +C h N h S h ;C2=γC1; Where: C1 is the system equipment investment cost; C2 is the system equipment operation and maintenance cost; C3 is the coal consumption and start-up and shutdown cost of the thermal power unit; C s , C pv , C w and C h are the unit capacity investment costs of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; N s 、N pv 、N w and N h are the number of hydropower units, photovoltaic arrays, wind power units and thermal power units respectively; S s , S pv , S w and S h are the total capacities of hydropower units, photovoltaic arrays, wind turbines and thermal power units respectively; γ is the operation and maintenance cost coefficient; C mh is the coal consumption cost of thermal power units; C q is the start-up and shutdown cost of the thermal power unit; T is the unit output dispatch cycle; P h (t) is the output power of a single thermal power unit at time t; a, b, c are the consumption characteristic coefficients of the thermal power unit; u t A is the start and stop status of the thermal power unit; h is the start-up and shutdown cost of thermal power units; ③ Determine the minimum net load fluctuation sub-goal, which is expressed as: Where: Pgh is the net load value at time t; Pgh,av is the average net load value within a dispatching cycle; ④ Based on seasonal factors, consider the maximum sub-goal of renewable energy power generation in three situations: normal, high and low water seasons: During the flood season, hydropower is fully generated, and its sub-objective expression is: During the dry season, the wind and solar power generation is the largest, and its sub-objective expression is: During the flat water period, the wind and solar power generation is the largest, and its sub-objective expression is: Where: T is the unit output dispatch cycle; P s (t), P pv (t) and P w (t) are the output power of a single hydropower unit, a unit photovoltaic array, and a single wind turbine unit at time t; ⑤ Determine the total carbon emission quantum target, the expression is: max f4=λE C Where: λ is the carbon emission factor; E C is the total thermal power generation; ⑥ Perform Min-Max normalization on the four sub-targets f1, f2, f3, and f4 respectively: ⑦ Perform linear weighted summation on the sub-goals and establish a multi-energy complementary system optimization model: min Z′=λ11f1'+λ2f'2+λ3f3'+λ4f4' Where: f1', f2', f3' and f4' are the normalized values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1max 、f 2max 、f 3max 、f 4max are the maximum values ​​of investment cost limit, load power shortage rate, renewable energy power generation and total carbon emissions respectively; f 1min 、f 2min 、f 3min 、f 4min They are respectively the minimum values ​​of investment cost limit, load power shortage rate, new energy power generation and total carbon emissions.

7. The capacity optimization configuration method of a seasonal water, fire, wind and solar multi-energy complementary system according to claim 1, characterized in that: The constraint conditions in step S3 are established as follows: ⑴ System single machine installed capacity constraints: Where: Z s , Z h , Z w , Z pv are the installed capacity of hydropower, thermal power, wind power and photovoltaic units respectively; Z smin , Z hmin , Z pvmin , Z wmin are the lower limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power generation and wind power generation in the target year; Z smax , Z hmax , Z pvmax , Z wmax They are the upper limits of the pre-planned installed capacity of hydropower, thermal power, photovoltaic power and wind power in the target year; ⑵Energy waste rate constraint: Where: D WE is the energy waste rate; D WEmax is the maximum energy waste rate; P WE (t) is the energy wasted power at time t; E is the total load power consumption within T time; ⑶ Thermal power safety startup restrictions: H≥C hmin R hmin T max Where: C hmin is the thermal power safety startup capacity obtained according to the regional power grid safety and stability constraints; R hmin is the minimum output rate of thermal power; T max is the number of hours in a year, 8760 hours in a non-leap year and 8784 hours in a leap year; (4) Output constraints of hydroelectric generators, photovoltaic generators, wind generators and thermal power generation: Where: P smin is the minimum output power of the unit hydropower unit; P smax , P pvmax and P wmax is the maximum output power of hydropower, photovoltaic and wind turbine units; P h is the output power of the thermal power unit; H min , H max It is the minimum output coefficient and maximum output coefficient of thermal power units.

Images