Nonlinear load prediction optimization method and system for hydropower engineering

Abstract

The invention relates to a nonlinear load prediction optimization method and system for hydropower engineering. The method comprises the following steps: collecting historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower engineering; historical hydroelectric load data, power grid demand data, meteorological data and reservoir operation data of the hydroelectric engineering are preprocessed to construct a load comprehensive data set; constructing a non-linear load prediction model, training the non-linear load prediction model based on the load comprehensive data set, and predicting the load of the hydropower engineering through the trained non-linear load prediction model; and constructing a multi-objective optimization function, and optimizing the predicted hydropower engineering load through the multi-objective optimization function to obtain an optimized hydropower engineering load.

Description

Technical Field

[0001] The invention relates to a nonlinear load prediction optimization method and system for a hydropower project, belonging to the technical field of hydropower projects. Background Art

[0002] In the field of hydropower load forecasting, some traditional methods, such as simple linear regression and ARIMA-type time series models, usually assume that there is a linear or approximately linear relationship between input and output; however, hydropower load is affected by many factors, such as meteorology, water level, unit scheduling decisions, grid electricity demand, seasonality, etc. These factors often have highly nonlinear coupling and time-lag characteristics. Linear models are difficult to capture the above complex interaction process, resulting in limited prediction accuracy. Summary of the invention

[0003] In order to solve the above problems existing in the prior art, the present invention proposes a nonlinear load prediction optimization method and system for hydropower projects.

[0004] The technical solution of the present invention is as follows:

[0005] In one aspect, the present invention provides a nonlinear load forecasting optimization method for a hydropower project, comprising the following steps:

[0006] Collect historical hydropower load data, grid demand data, meteorological data, and reservoir operation data of hydropower projects;

[0007] The historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project are preprocessed to construct a comprehensive load data set;

[0008] Construct a nonlinear load prediction model, train the nonlinear load prediction model based on the comprehensive load data set, and predict the load of the hydropower project through the trained nonlinear load prediction model;

[0009] A multi-objective optimization function is constructed, and the predicted hydropower project load is optimized through the multi-objective optimization function to obtain the optimized hydropower project load.

[0010] As a preferred embodiment of the present invention, the specific steps of preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project include: missing value filling, outlier detection and data smoothing.

[0011] As a preferred embodiment of the present invention, the nonlinear load forecasting model is constructed based on a neural network, as shown in the following formula:

[0012] L p (t)=Res(t)+Neu(t)+W(t)+ε(t)

[0013] Where: L p (t) represents the predicted load of the hydropower project at time t; Res(t) represents the reservoir water level and flow coupling polynomial; Neu(t) represents the neural network activation function; W(t) represents the meteorological time function; ε(t) represents random noise.

[0014] As a preferred embodiment of the present invention, the reservoir water level flow coupling polynomial is specifically shown as follows:

[0015]

[0016] Where: K represents the upper limit of the sum of the outer layer of the polynomial; α k represents the weight coefficient of the kth group of polynomial combinations; P represents the highest power of the polynomial; β kp represents the coefficient of the pth power of the kth group in the polynomial; Q(t-τ kp ) represents t-τ kp The inflow and outflow of the reservoir at the moment; τ kp represents the delay step length of the reservoir outflow; η kp Represents the reservoir flow-water level coupling coefficient; H(t-ζ kp ) represents t-ζ kp The reservoir water level at the moment; kp Indicates the delay step of the reservoir water level.

[0017] As a preferred embodiment of the present invention, the neural network activation function is specifically shown in the following formula:

[0018]

[0019] Where: M represents the number of neurons in the neural network; ω m represents the activation output weight of the mth neuron; σ represents the activation function of the neural network; γ m represents the scaling factor of the activation function of the mth neuron; N represents the number of input samples; θ mn represents the weight of the mth neuron activation function corresponding to the nth input sample feature; X n (td mn ) indicates that the nth input sample feature is in td mn The value of the moment; d mn represents the delay time of the mth neuron at the nth input sample; b m Represents the bias of the activation function of the mth neuron.

[0020] As a preferred embodiment of the present invention, the meteorological time function is specifically shown in the following formula:

[0021]

[0022] Where: R represents the number of exponential components; δ r represents the amplitude factor of the rth exponential component; T w (tr) represents the meteorological data sample at time tr; μ r represents the attenuation coefficient of the rth exponential component.

[0023] As a preferred embodiment of the present invention, a multi-objective optimization function J is constructed to optimize the predicted hydropower project load, as shown in the following formula:

[0024]

[0025] Where: L r (t) represents the actual load of the hydropower project at time t; H * (t) represents the expected water level of the reservoir at time t; H(t) represents the actual water level of the reservoir at time t; Q spill (t) represents the overflow of the reservoir at time t; P gen (t) represents the power generation of the hydropower station unit at time t; P ref (t) represents the reference demand of the load side at time t; S(t) represents the penalty term; α, β, γ, and δ represent the weight coefficients respectively.

[0026] As a preferred implementation of the present invention, the penalty term is specifically shown in the following formula:

[0027]

[0028] Where: H max Indicates the upper limit of the reservoir water level; H min represents the lower limit of reservoir water level; κ and λ represent weights.

[0029] As a preferred embodiment of the present invention, the multi-objective optimization function is provided with the following constraints:

[0030] Water balance constraints:

[0031]

[0032] Where: H(t+1) represents the reservoir water level at time t+1; A(t) represents the cross-sectional area of ​​the reservoir water level; Q in (t) represents the inflow of the reservoir at time t; Q out (t) represents the outflow of the reservoir at time t; Δt represents the time step;

[0033] Power generation constraints:

[0034] P gen (t) = ρgη turb (t)Qturb (t)[H res (t)-H tail (t)]

[0035] Where: ρ represents water density; g represents gravitational acceleration; η turb (t) represents the power generation efficiency of the hydropower station unit at time t; Q turb (t) represents the permeation flow rate of the hydropower station unit at time t; H res (t) represents the upstream water head at time t; H tail (t) represents the tail water level at time t.

[0036] On the other hand, the present invention also provides a nonlinear load forecasting and optimization system for hydropower engineering, comprising a data acquisition module, a data processing module, a load forecasting module and a forecasting and optimization module;

[0037] The data acquisition module is used to collect historical hydropower load data, grid demand data, meteorological data and reservoir operation data of the hydropower project;

[0038] The data processing module is used to construct a comprehensive load data set after preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project;

[0039] The load prediction module is used to construct a nonlinear load prediction model, train the nonlinear load prediction model based on the load comprehensive data set, and predict the hydropower project load through the trained nonlinear load prediction model;

[0040] The prediction optimization module is used to construct a multi-objective optimization function, and optimize the predicted hydropower project load through the multi-objective optimization function to obtain the optimized hydropower project load.

[0041] The present invention has the following beneficial effects:

[0042] 1. The present invention describes the nonlinear coupling of flow and water level under different time lags through polynomial terms, thus overcoming the complex relationship between water level, flow and unit efficiency that is difficult to handle with traditional linear models.

[0043] 2. Based on the prediction, the present invention further constructs a multi-objective optimization function, taking into account multiple goals such as prediction accuracy, water level stability, minimization of water abandonment, unit power generation deviation, and water level upper and lower limit safety. It can better balance multiple demands such as power generation efficiency, flood control safety, and ecological discharge, rather than just focusing on a single performance indicator. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be understood that the step numbers used in this document are only for convenience of description and are not intended to limit the order in which the steps are executed.

[0047] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0048] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0049] The term "and / or" means and includes any and all possible combinations of one or more of the associated listed items.

[0050] Embodiment 1:

[0051] See also Figure 1 , a nonlinear load forecasting optimization method for a hydropower project, comprising the following steps:

[0052] Collect historical hydropower load data, grid demand data, meteorological data, and reservoir operation data of hydropower projects;

[0053] The historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project are preprocessed to construct a comprehensive load data set;

[0054] Construct a nonlinear load prediction model, train the nonlinear load prediction model based on the comprehensive load data set, and predict the load of the hydropower project through the trained nonlinear load prediction model;

[0055] A multi-objective optimization function is constructed, and the predicted hydropower project load is optimized through the multi-objective optimization function to obtain the optimized hydropower project load.

[0056] As a preferred implementation of this embodiment, the specific steps of preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project include: missing value filling, outlier detection and data smoothing.

[0057] As a preferred implementation of this embodiment, the nonlinear load forecasting model is constructed based on a neural network, as shown in the following formula:

[0058] L p (t)=Res(t)+Neu(t)+W(t)+ε(t)

[0059] Where: L p (t) represents the predicted load of the hydropower project at time t; Res(t) represents the reservoir water level and flow coupling polynomial; Neu(t) represents the neural network activation function; W(t) represents the meteorological time function; ε(t) represents random noise.

[0060] As a preferred implementation of this embodiment, the load (or power generation) of a hydropower station often has a complex physical relationship with the upstream and downstream flow and the water level in the reservoir area. However, in actual practice, this relationship is not a simple linear relationship: there are many factors such as the unit efficiency curve, head-flow coupling, and scheduling delay;

[0061] Expanding the flow and water level in a polynomial form can capture this complex nonlinear relationship more flexibly. The reservoir water level flow coupling polynomial is specifically shown in the following formula:

[0062]

[0063] Among them: K represents the upper limit of the sum of the outer layer of the polynomial. The larger K is, the more polynomial combinations there are; α k Represents the weight coefficient of the kth group of polynomial combinations, which is used to regulate the contribution of this group of polynomials to the overall prediction results; P represents the highest power of the polynomial. The larger the power, the higher the polynomial order, the stronger the nonlinear fitting ability, but it is easy to overfit; β kp represents the coefficient of the pth power of the kth group in the polynomial; Q(t-τ kp ) represents t-τ kp The inflow and outflow of the reservoir at the moment represents the time lag effect, that is, the impact of the flow at several moments in the past on the current load; τ kp represents the delay step length of the reservoir outflow; η kp Represents the reservoir flow-water level coupling coefficient; H(t-ζ kp ) represents t-ζ kp The reservoir water level at the moment is similar to the flow rate. The time lag state of the water level in the past has an impact on the current load; kp Indicates the delay step of the reservoir water level.

[0064] As a preferred implementation of this embodiment, the neural network activation function is specifically shown in the following formula:

[0065]

[0066] Where: M represents the number of neurons in the neural network; ω m represents the activation output weight of the mth neuron; σ represents the activation function of the neural network, such as sigmoid, tanh, ReLU, etc.; γ m represents the scaling factor of the activation function of the mth neuron; N represents the number of input samples; θ mn represents the weight of the mth neuron activation function corresponding to the nth input sample feature; X n (td mn ) indicates that the nth input sample feature is in td mn The value of the moment; d mn represents the delay time of the mth neuron at the nth input sample; b m Represents the bias of the activation function of the mth neuron.

[0067] As a preferred implementation of this embodiment, the meteorological time function is specifically shown in the following formula:

[0068]

[0069] Among them: R represents the number of index components, which controls the number of meteorological related items. The more the number, the more complex the meteorological influence forms can be fitted; δ r represents the amplitude factor of the rth exponential component, which controls the contribution intensity of the exponential kernel in the overall prediction result; T w (tr) represents the meteorological data sample (such as precipitation, humidity, etc.) at time tr; μ r Represents the decay coefficient of the rth exponential component, which controls the "width" of the Gaussian distribution and determines how the meteorological impact changes with temperature (or other meteorological variables).

[0070] As a preferred implementation of this embodiment, a multi-objective optimization function J is constructed to optimize the predicted hydropower project load, as shown in the following formula:

[0071]

[0072] Where: L r (t) represents the actual load of the hydropower project at time t; H * (t) represents the expected water level of the reservoir at time t; H(t) represents the actual water level of the reservoir at time t; Q spill (t) represents the overflow of the reservoir at time t; P gen(t) represents the power generation of the hydropower station unit at time y; P ref (t) represents the reference demand of the load side at time t; S(t) represents the penalty term; α, β, γ, and δ represent weight coefficients respectively;

[0073] In this embodiment, the objective function is solved by a hybrid heuristic algorithm, and the specific steps are as follows:

[0074] The initial solution obtains multiple groups of populations through methods such as Latin square sampling (LHS) or Sobol sequence;

[0075] Set custom crossover, mutation and fitness functions for load forecasting models and scheduling equations;

[0076] After using genetic algorithm for global search, particle swarm optimization is used for local search to improve the convergence speed;

[0077] Finally, gradient descent or Newton's method is used for fine-tuning in the convergence region to further improve the accuracy.

[0078] As a preferred implementation of this embodiment, the penalty term is specifically shown in the following formula:

[0079]

[0080] Where: H max Represents the upper limit of the reservoir water level. If the water level exceeds the upper limit, a penalty will be incurred to ensure flood control safety; H min represents the lower limit of the reservoir water level. If the water level is lower than the lower limit, a penalty will be incurred to ensure water supply and ecological needs; κ and λ represent weights.

[0081] As a preferred implementation of this embodiment, the multi-objective optimization function is provided with the following constraints:

[0082] Water balance constraints:

[0083]

[0084] Where: H(t+1) represents the reservoir water level at time t+1; A(t) represents the cross-sectional area of ​​the reservoir water level (which can be expressed approximately or in sections); Q in (t) represents the inflow of the reservoir at time t; Q out (t) represents the outflow of the reservoir at time t; Δt represents the time step;

[0085] Power generation constraints:

[0086] P gen (t) = ρgη turb (t)Q turb (t)[H res (t)-Htail (t)]

[0087] Where: ρ represents water density; g represents gravitational acceleration; η turb (t) represents the power generation efficiency of the hydropower station unit at time t; Q turb (t) represents the permeation flow rate of the hydropower station unit at time t; H res (t) represents the upstream water head at time t; H tail (t) represents the tail water level at time t.

[0088] Embodiment 2:

[0089] A nonlinear load forecasting and optimization system for a hydropower project, comprising a data acquisition module, a data processing module, a load forecasting module and a forecasting and optimization module;

[0090] The data acquisition module is used to collect historical hydropower load data, grid demand data, meteorological data and reservoir operation data of the hydropower project;

[0091] The data processing module is used to construct a comprehensive load data set after preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project;

[0092] The load prediction module is used to construct a nonlinear load prediction model, train the nonlinear load prediction model based on the load comprehensive data set, and predict the hydropower project load through the trained nonlinear load prediction model;

[0093] The prediction optimization module is used to construct a multi-objective optimization function, and optimize the predicted hydropower project load through the multi-objective optimization function to obtain the optimized hydropower project load.

[0094] The system is used to implement the method in Example 1, which will not be described in detail here.

[0095] Embodiment three:

[0096] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described in any embodiment of the present invention when executing the program.

[0097] Embodiment 4:

[0098] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method described in any embodiment of the present invention is implemented.

[0099] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c can be single or multiple.

[0100] Those of ordinary skill in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented in a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0101] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0102] In several embodiments provided in the present application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory; hereinafter referred to as: ROM), random access memory (Random Access Memory; hereinafter referred to as: RAM), disk or optical disk, and other media that can store program codes.

[0103] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A nonlinear load forecasting optimization method for a hydropower project, characterized in that: The following steps are involved: Collect historical hydropower load data, grid demand data, meteorological data, and reservoir operation data of hydropower projects; The historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project are preprocessed to construct a comprehensive load data set; Construct a nonlinear load prediction model, train the nonlinear load prediction model based on the comprehensive load data set, and predict the load of the hydropower project through the trained nonlinear load prediction model; A multi-objective optimization function is constructed, and the predicted hydropower project load is optimized through the multi-objective optimization function to obtain the optimized hydropower project load.

2. A nonlinear load forecasting optimization method for a hydropower project according to claim 1, characterized in that: The specific steps of preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project include: missing value filling, outlier detection and data smoothing.

3. The nonlinear load prediction optimization method for a hydropower project according to claim 1, characterized in that: The nonlinear load forecasting model is constructed based on a neural network, as shown in the following formula: L p (t)=Res(t)+Neu(t)+W(t)+ε(t) Where: L p (t) represents the predicted load of the hydropower project at time t; Res(t) represents the reservoir water level and flow coupling polynomial; Neu(t) represents the neural network activation function; W(t) represents the meteorological time function; ε(t) represents random noise.

4. A nonlinear load forecasting optimization method for a hydropower project according to claim 3, characterized in that: The reservoir water level-flow coupling polynomial is specifically shown in the following formula: Where: K represents the upper limit of the outer summation of the polynomial; α k represents the weight coefficient of the kth group of polynomial combinations; P represents the highest power of the polynomial; β kp represents the coefficient of the pth power of the kth group in the polynomial; Q(t-τ kp ) represents t-τ kp The inflow and outflow of the reservoir at the moment; τ kp represents the delay step length of the reservoir outflow; η kp represents the reservoir flow-water level coupling coefficient; express Reservoir water level at the moment; Indicates the delay step of the reservoir water level.

5. The nonlinear load forecasting optimization method for a hydropower project according to claim 3, characterized in that: The neural network activation function is specifically shown in the following formula: Where: M represents the number of neurons in the neural network; ω m represents the activation output weight of the mth neuron; σ represents the activation function of the neural network; γ m represents the scaling factor of the activation function of the mth neuron; N represents the number of input samples; θ mn represents the weight of the mth neuron activation function corresponding to the nth input sample feature; X n (td mn ) indicates that the nth input sample feature is in td mn The value of the moment; d mn represents the delay time of the mth neuron at the nth input sample; b m Represents the bias of the activation function of the mth neuron.

6. The nonlinear load forecasting optimization method for a hydropower project according to claim 3, characterized in that: The meteorological time function is specifically shown in the following formula: Where: R represents the number of exponential components; δ r represents the amplitude factor of the rth exponential component; T w (tr) represents the meteorological data sample at time tr; μ r represents the attenuation coefficient of the rth exponential component.

7. The nonlinear load forecasting optimization method for a hydropower project according to claim 3 is characterized in that: A multi-objective optimization function J is constructed to optimize the predicted hydropower project load, as shown in the following formula: Where: L r (t) represents the actual load of the hydropower project at time t; H * (t) represents the expected water level of the reservoir at time t; H(t) represents the actual water level of the reservoir at time t; Q spill (t) represents the overflow of the reservoir at time t; P gen (t) represents the power generation of the hydropower station unit at time t; P ref (t) represents the reference demand of the load side at time t; S(t) represents the penalty term; α, β, γ, and δ represent the weight coefficients respectively.

8. A nonlinear load forecasting optimization method for a hydropower project according to claim 7, characterized in that: The penalty term is specifically shown in the following formula: Where: H max Indicates the upper limit of the reservoir water level; H min represents the lower limit of reservoir water level; κ and λ represent weights.

9. A nonlinear load forecasting optimization method for a hydropower project according to claim 7, characterized in that: The multi-objective optimization function is subject to the following constraints: Water balance constraints: Where: H(t+1) represents the reservoir water level at time t+1; A(t) represents the cross-sectional area of ​​the reservoir water level; Q in (t) represents the inflow of the reservoir at time t; Q out (t) represents the outflow of the reservoir at time t; Δt represents the time step; Power generation constraints: P gen (t)=ρgη turb (t)Q turb (t)[H res (t)-H tail (t)] Where: ρ represents water density; g represents gravitational acceleration; η turb (t) represents the power generation efficiency of the hydropower station unit at time t; Q turb (t) represents the permeation flow rate of the hydropower station unit at time t; H res (t) represents the upstream water head at time t; H tail (t) represents the tail water level at time t.

10. A nonlinear load forecasting optimization system for hydropower projects, characterized in that: It includes data acquisition module, data processing module, load forecasting module and forecasting optimization module; The data acquisition module is used to collect historical hydropower load data, grid demand data, meteorological data and reservoir operation data of the hydropower project; The data processing module is used to construct a comprehensive load data set after preprocessing the historical hydropower load data, power grid demand data, meteorological data and reservoir operation data of the hydropower project; The load prediction module is used to construct a nonlinear load prediction model, train the nonlinear load prediction model based on the load comprehensive data set, and predict the hydropower project load through the trained nonlinear load prediction model; The prediction optimization module is used to construct a multi-objective optimization function, and optimize the predicted hydropower project load through the multi-objective optimization function to obtain the optimized hydropower project load.

Images