Water balance regulation and control method and device of power plant, electronic equipment and storage medium

Abstract

The invention relates to a water balance regulation and control method and device for a power plant, electronic equipment and a storage medium, and the method comprises the steps: inputting the multi-source operation data of a current power plant water system into a preset time sequence prediction model, outputting a water quantity prediction result, and based on a preset multi-target optimization algorithm, according to the current unit operation state and the water quantity prediction result, obtaining a water balance regulation and control result; obtaining a target control strategy meeting a preset constraint condition; inputting the current wastewater quality and quantity information into a preset wastewater scheduling model for wastewater utilization planning, and generating a target wastewater utilization strategy; and generating a target control instruction according to the target control strategy and the target wastewater utilization strategy, and issuing the target control instruction to an execution mechanism so as to control the execution mechanism to perform a corresponding adjustment action based on the target control instruction. Therefore, the static, isolated and extensive defects of existing coal-fired power plant water system management are overcome, the unit power generation water taking amount is remarkably reduced, the comprehensive utilization efficiency of water resources is improved, and the method adapts to the complex operation working conditions such as deep peak regulation.

Description

Technical Field

[0001] This application relates to the field of smart power plant technology, and in particular to a water balance control method, device, electronic equipment and storage medium for a power plant. Background Technology

[0002] As major industrial water users, coal-fired power plants have complex water systems that include multiple subsystems such as raw water pretreatment, boiler feedwater, circulating cooling water, industrial wastewater, and desulfurization wastewater.

[0003] Currently, the industry's management of plant-wide water balance mainly relies on manual experience and periodic water balance tests, which has obvious problems of static, isolated, and extensive management.

[0004] Traditional methods cannot respond in real time to dynamic changes in unit load, ambient temperature, and water quality fluctuations. Furthermore, each water subsystem often operates independently, lacking a global collaborative scheduling and optimization mechanism, resulting in low water reuse rates and persistently high fresh water consumption. Simultaneously, existing water-saving measures often focus on a single water-saving objective, neglecting the coupling relationship with multiple objectives such as system energy consumption, equipment safety, and environmental emissions. This makes it difficult to achieve an optimal balance between economic and environmental benefits, a problem that urgently needs to be addressed. Summary of the Invention

[0005] This application provides a water balance control method, device, electronic equipment, and storage medium for power plants to overcome the static, isolated, and extensive defects of existing coal-fired power plant water system management, thereby significantly reducing the water intake per unit of power generation, improving the comprehensive utilization efficiency of water resources, and adapting to complex operating conditions such as deep peak shaving.

[0006] The first aspect of this application provides a water balance control method for a power plant, characterized by comprising the following steps: Obtain the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system; The multi-source operating data is input into a preset time-series prediction model, and the water volume prediction result is output through the preset time-series prediction model. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and the water volume prediction result. The current wastewater quality and quantity information is input into a preset wastewater scheduling model, and wastewater utilization planning is carried out through the preset wastewater scheduling model to generate a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy, and then sent to the actuator to control the actuator to perform corresponding adjustment actions based on the target control instructions.

[0007] According to one embodiment of this application, the objective function of the preset multi-objective optimization algorithm is: F(X) = W 1 f 1 +W 2 f 2 +W 3 f 3; f1=Q fresh / P ; ; ; Where F(X) is the objective function, f 1 represents the water intake per unit of power generation. f 2 represents auxiliary power consumption of the water system. f 3 represents the pollution load of discharged wastewater. Q fresh This refers to the amount of fresh water taken. P For the power generation of the unit, Let be the power of the i-th pump. For the density of water, It is the acceleration due to gravity. Let i be the flow rate of the i-th pump. Let i be the head of the i-th pump. Let the efficiency of the i-th pump be... For motor efficiency, For the j-th discharge wastewater flow rate, Let n be the concentration of the nth pollutant in the j-th wastewater. W 1. W 2. W 3 are all weighting coefficients.

[0008] According to one embodiment of this application, the scheduling objective function of the preset wastewater scheduling model is: ; in, Q i,j Let i be the amount of water transported from water source i to water point j. L i,j Let λ be the water transport distance from water source i to water point j, and λ be the energy consumption coefficient per unit distance. β For the reuse rate weight, ∑ Q source,i This represents the total wastewater volume.

[0009] According to one embodiment of this application, after obtaining the current unit operating status of the power plant, the current wastewater quality and quantity information, and the multi-source operating data of the water system, the method further includes: Monitor the operating status of the water system; When the operating status of the water system meets the preset risk conditions, an early warning reminder is generated, and a risk warning reminder is issued based on the preset digital twin platform; The early warning information includes at least one of fault location information and risk handling strategies.

[0010] According to one embodiment of this application, before inputting the multi-source operational data into a preset time-series prediction model, the method further includes: The multi-source operational data is cleaned to obtain cleaned multi-source operational data; The cleaned multi-source operating data is subjected to outlier removal processing to obtain processed multi-source operating data; The processed multi-source operational data is standardized and converted, and the standardized operational data is stored in a preset time-series database.

[0011] According to one embodiment of this application, the multi-source operating data includes at least one of flow rate, pressure, temperature, pH value, conductivity, turbidity, hardness, and chloride ion content, and the water volume prediction results include at least one of water demand prediction, water production prediction, and reclaimable water volume prediction.

[0012] According to the water balance control method for power plants provided in this application, the multi-source operation data of the current power plant water system is input into a preset time-series prediction model, which outputs water volume prediction results. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and water volume prediction results. The current wastewater quality and quantity information is input into a preset wastewater scheduling model for wastewater utilization planning, generating a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy and sent to the actuators to control the actuators to perform corresponding adjustment actions based on the target control instructions. This overcomes the static, isolated, and extensive defects of existing coal-fired power plant water system management, significantly reducing water intake per unit of power generation, improving the comprehensive utilization efficiency of water resources, and adapting to complex operating conditions such as deep peak shaving.

[0013] A second aspect of this application provides a water balance control device for a power plant, comprising: The acquisition module is used to acquire the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system. The processing module is used to input the multi-source operating data into a preset time-series prediction model, output water volume prediction results through the preset time-series prediction model, and obtain a target control strategy that meets preset constraints based on a preset multi-objective optimization algorithm, according to the current unit operating status and the water volume prediction results. The wastewater scheduling module is used to input the current wastewater quality and quantity information into a preset wastewater scheduling model, and to perform wastewater utilization planning through the preset wastewater scheduling model to generate a target wastewater utilization strategy. The control module is used to generate target control instructions based on the target control strategy and the target wastewater utilization strategy, and to send the target control instructions to the actuator to control the actuator to perform corresponding adjustment actions based on the target control instructions.

[0014] According to one embodiment of this application, the objective function of the preset multi-objective optimization algorithm is: F(X) = W 1 f 1 +W 2 f 2 +W 3 f 3; f1=Q fresh / P ; ; ; Where F(X) is the objective function, f 1 represents the water intake per unit of power generation. f 2 represents auxiliary power consumption. f 3 represents the pollution load of discharged wastewater. Q fresh This refers to the amount of fresh water taken. P For the power generation of the unit, Let be the power of the i-th pump. For the density of water, It is the acceleration due to gravity. Let i be the flow rate of the i-th pump. Let i be the head of the i-th pump. Let the efficiency of the i-th pump be... For motor efficiency, For the j-th discharge wastewater flow rate, Let n be the concentration of the nth pollutant in the j-th wastewater. W 1. W 2. W 3 are all weighting coefficients.

[0015] According to one embodiment of this application, the scheduling objective function of the preset wastewater scheduling model is: ; in, Q i,j Let i be the amount of water transported from water source i to water point j. L i,jLet λ be the water transport distance from water source i to water point j, and λ be the energy consumption coefficient per unit distance. β For the reuse rate weight, ∑ Q source,i This represents the total wastewater volume.

[0016] According to one embodiment of this application, after acquiring the current unit operating status of the power plant, the current wastewater quality and quantity information, and the multi-source operating data of the water system, the acquisition module is further configured to: Monitor the operating status of the water system; When the operating status of the water system meets the preset risk conditions, an early warning reminder is generated, and a risk warning reminder is issued based on the preset digital twin platform; The early warning information includes at least one of fault location information and risk handling strategies.

[0017] According to one embodiment of this application, before inputting the multi-source operational data into a preset time-series prediction model, the processing module is further configured to: The multi-source operational data is cleaned to obtain cleaned multi-source operational data; The cleaned multi-source operating data is subjected to outlier removal processing to obtain processed multi-source operating data; The processed multi-source operational data is standardized and converted, and the standardized operational data is stored in a preset time-series database.

[0018] According to one embodiment of this application, the multi-source operating data includes at least one of flow rate, pressure, temperature, pH value, conductivity, turbidity, hardness, and chloride ion content, and the water volume prediction results include at least one of water demand prediction, water production prediction, and reclaimable water volume prediction.

[0019] According to the water balance control device for power plants provided in this application embodiment, the multi-source operation data of the current power plant water system is input into a preset time-series prediction model, outputting water volume prediction results. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and water volume prediction results. The current wastewater quality and quantity information is input into a preset wastewater scheduling model for wastewater utilization planning, generating a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy and sent to the actuators to control the actuators to perform corresponding adjustment actions based on the target control instructions. This overcomes the static, isolated, and extensive defects of existing coal-fired power plant water system management, significantly reducing water intake per unit of power generation, improving the comprehensive utilization efficiency of water resources, and adapting to complex operating conditions such as deep peak shaving.

[0020] A third aspect of this application 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 executes the program to implement the power plant water balance control method as described in the above embodiments.

[0021] A fourth aspect of this application provides a computer-readable storage medium storing computer instructions for causing the computer to perform the power plant water balance control method as described in the above embodiments.

[0022] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a water balance control method for a power plant according to an embodiment of this application; Figure 2 This is a block diagram of a water balance control device for a power plant according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0025] Those skilled in the art will understand that, currently, the industry generally faces the following technical bottlenecks in the management of overall plant water balance: (1) Static management: Traditional water balance tests rely on manual labor and are conducted periodically (usually once every 3-5 years). They cannot respond in real time to the dynamic changes in unit load, ambient temperature, and water quality fluctuations. The results are delayed and have limited guiding significance for daily refined operation.

[0026] (2) System isolation: Each water subsystem (such as circulating water, industrial wastewater treatment, and desulfurization water) often operates and is controlled independently, lacking coordinated scheduling and overall optimization of water resources throughout the plant. A large number of secondary and tertiary water systems lack metering, resulting in a "black box" of water flow, low reuse rate, and high consumption of fresh water.

[0027] (3) Inefficient control: Key operations such as water replenishment, sewage discharge, and chemical dosing rely heavily on the experience of operators or simple PID control, making it impossible to achieve prediction-based feedforward optimization. For critical water volumes such as evaporation loss, wind loss, and abnormal leakage, there is a lack of accurate online calculation and diagnostic models.

[0028] (4) Optimization of single targets: Existing water-saving measures focus on single targets (such as increasing the concentration ratio) and ignore the mutual constraints between system energy consumption (such as fan and pump consumption), equipment scaling risk, environmental emissions and other targets, making it difficult to achieve the optimal balance between economic benefits and environmental benefits.

[0029] (5) Although some standards have put forward requirements for testing and some smart water platforms have realized data monitoring, there is a lack of an integrated solution that can realize plant-wide scale, dynamic prediction, multi-objective collaborative optimization and intelligent closed-loop control.

[0030] Therefore, there is an urgent need for a dynamic water balance control technology capable of real-time sensing, intelligent decision-making, and precise execution. Based on this, this application proposes a dynamic water balance intelligent control system and method for coal-fired power plants, based on multi-source data sensing, deep learning prediction, multi-objective intelligent optimization, and digital twin visualization. This system is suitable for the efficient utilization and precise dynamic balance control of water resources in coal-fired power plants under complex operating conditions such as deep peak shaving.

[0031] Before introducing the power plant water balance control method of this application, let's first introduce the multi-dimensional coupled intelligent control system for dynamic water balance of coal-fired power plants proposed in this application.

[0032] This multi-dimensional coupled intelligent control system for dynamic water balance in coal-fired power plants is based on a five-layer architecture of "sensing-computation-decision-visualization-execution," constructing a closed-loop control system for the entire process, specifically including: (1) Multi-source data sensing layer, consisting of online monitoring instruments for flow, pressure, temperature, and water quality monitoring parameters (including pH, conductivity, turbidity, hardness, chloride ion content, etc.) deployed at key nodes of the plant's water system, as well as the unit's DCS / PLC / MIS system and environmental monitoring system, to realize real-time acquisition of process data from no less than 500 measuring points covering three major types of raw water, circulating water, and wastewater, including raw water pretreatment, boiler feedwater, industrial water, fire water, circulating cooling water, closed cooling water, coal-containing wastewater, industrial wastewater, and desulfurization wastewater, providing a comprehensive and accurate data source for system control.

[0033] Optionally, in this embodiment of the application, electromagnetic flowmeters (measurement range 0-500 m³ / h), pressure transmitters (measurement range 0-1.6 MPa), pH online monitors (measurement range 0-14), conductivity meters (measurement range 0-20000 μS / cm), and hardness online monitors can be deployed at key nodes such as raw water inlet, circulating water cooling tower, desulfurization wastewater treatment station, and industrial wastewater reuse point to achieve full-process data acquisition with a acquisition frequency of 1 second / time.

[0034] (2) Edge computing and data hub layer, including time series database (plant-level real-time database, historical database) data preprocessing module. The data preprocessing module integrates data cleaning, outlier removal, standardization conversion and feature engineering functions. After noise reduction and normalization of the collected raw data, it is stored in the time series database. At the same time, it completes the computational tasks with high real-time requirements through edge computing.

[0035] Among them, the time-series database of the edge computing and data hub layer supports millisecond-level data writing and querying, meeting the real-time control requirements of the system.

[0036] (3) The core layer of intelligent decision-making is the core control software platform of the system. It is constructed by the software platform and includes three major functional modules: dynamic prediction module (Bi-LSTM-attention composite deep learning model), multi-objective optimization module (dynamic weight MOPSO algorithm), and wastewater dynamic scheduling module (colored Petri net (CPN) model).

[0037] 1) The dynamic prediction module adopts a composite deep learning model that combines a bidirectional long short-term memory network (Bi-LSTM) with an attention mechanism. The input is time series data such as historical and real-time unit load, environmental parameters, water quality and quantity data, and the output is the predicted value of water demand, water production and reclaimable water volume of each subsystem in the next 30-60 minutes, and the average absolute percentage error of the prediction is ≤5%.

[0038] The deep learning model employs a bidirectional long short-term memory network to capture the "past-present-future" dependency relationship of water quantity and water quality time-series data, overcoming the limitation of traditional LSTM unidirectional propagation in fully utilizing subsequent time-series features. An attention mechanism is introduced to dynamically assign weights to time-series features at different times, enhancing the contribution (i.e., weight allocation) of key influencing factors such as sudden changes in unit load and fluctuations in raw water quality, thereby improving prediction accuracy under complex operating conditions. The model is structured in four levels: forward LSTM layer → backward LSTM layer → attention fusion layer → fully connected output layer. The time-series input sequence is defined as X = [X1, X2, ..., Xt...XT], where Xt ∈ Rd is the input feature vector at time t (containing d features such as unit load, ambient temperature, and raw water conductivity), and T is the time step.

[0039] Furthermore, the bidirectional long short-term memory network includes a forget gate, an input gate, a cell state update gate, and an output gate. The forget gate controls the proportion of historical data states retained; the specific gating unit formula is as follows: The input gate controls the proportion of the current data feature input. The specific gating unit formula is as follows: Cell state updates store long-term temporal information; the specific gating unit formula is as follows: The output gate controls the output ratio of the current cell state. The specific gating unit formula is as follows: In the hidden state, the specific gating unit formula is: .

[0040] In the above gating formula, σ is the Sigmoid activation function, tanh is the hyperbolic tangent activation function, ⊙ is the element-wise product, Wf, Wi, Wc, Wo are the weight matrices, bf, bi, bc, bo are the bias vectors, and ht 1 represents the hidden state in the previous moment, Ct 1 represents the cell state at the previous moment.

[0041] Furthermore, the bidirectional LSTM fusion formula includes the transition from the forward LSTM output to the backward LSTM output. Forward LSTM output (from t=1 to t=T): Inverse LSTM output (from t=T to t=1): Two-way fusion features: (⊕ represents vector concatenation).

[0042] Furthermore, the attention mechanism weights are calculated as scaled dot product attention, where the attention score (calculated by comparing the current time step with all other time steps) is: Attention weights (normalized relevance score): Attention fusion output (weighted summation to enhance key features): dbi is the bidirectional fusion feature dimension, and αt,k is the attention weight from time k to time t.

[0043] Furthermore, the final prediction output is mapped to a predicted value (water volume for the next 30-60 minutes) through a fully connected layer: Where τ is the prediction step size (30-60 minutes corresponds to 5-10 time steps). This represents the predicted water volume at time t+τ.

[0044] Furthermore, model performance constraints: the mean absolute percentage error (MAPE) of predictions ≤ 5%, i.e., MAPE = ≤5%, where yi is the actual water volume. For the predicted water volume, N is the sample size.

[0045] 2) Multi-objective optimization module: The MOPSO (Multi-objective Particle Swarm Optimization) algorithm is used to construct a three-objective optimization model with the goals of "minimum water consumption per unit of power generation, minimum power consumption of water system, and minimum external pollution load". The module innovatively introduces dynamic weight coefficients for operating conditions. Under rated load conditions, the focus is on water conservation objectives, and under deep peak shaving conditions, the focus is on system stability objectives. Non-dominated solutions are screened through Pareto dominance relationship, and the optimal control strategy is finally determined by linear weighted summation method.

[0046] Furthermore, the multi-objective optimization module employs a multi-objective particle swarm optimization algorithm, aiming to minimize water consumption per unit of power generation, the total power consumption of the water system, and the pollution load of discharged wastewater. Under multiple constraints, including circulating water concentration ratio, key water quality indicators, and equipment operating ranges, it performs online rolling optimization, generating a new optimal control strategy set every 10 seconds. The multi-objective optimization module sets dynamic weighting coefficients for different operating conditions, emphasizing reducing water consumption per unit of power generation under rated load conditions and prioritizing system operational stability under deep peak-shaving conditions.

[0047] 3) The wastewater dynamic scheduling module abstracts the wastewater scheduling system into a colored Petri net (CPN) model. Here, "Places" correspond to wastewater sources / use points / transmission pipelines, "Transitions" correspond to the start and stop actions of wastewater transfer pumps / valvees, and the color of "Tokens" indicates the wastewater quality category (e.g., purple = high-salinity wastewater, blue = medium-salinity wastewater, green = low-salinity wastewater). The number of tokens represents the wastewater volume. Through changes in the state of the places and the triggering rules of transitions, dynamic optimization of wastewater reuse paths is achieved.

[0048] The wastewater dynamic scheduling module's colored Petri net model supports real-time updates of wastewater source water quality parameters and water quality demand parameters at water usage points, adapting to dynamic changes in system operating conditions. In other words, the wastewater dynamic scheduling module establishes a scheduling model based on a colored Petri net, modeling various wastewater sources (such as reverse osmosis concentrate, chemical backwash water, and desulfurization wastewater) and their water quality attributes, as well as each water usage point (such as circulating water makeup, coal conveying flushing, and dry slag humidification) and its water quality requirements as a network. The color attribute of the tokens in the model represents different water quality categories. The scheduling engine dynamically triggers transitions based on real-time water quality and quantity data and water usage point requirements to guide wastewater flow to the optimal reuse path, and calculates the optimal wastewater allocation path and flow rate in real time.

[0049] Furthermore, key definitions and scheduling rules: ① Definition of core elements of the CPN model: The set of places P={Psource,1,...Psource,n,...Pwaterpoint,1,...,Pwater point,m,...,.Ppipe,1,...,Ppipe,k} corresponds to wastewater source, water point, and water pipeline respectively; a. The transition set T = {T1, T2, ..., Tl} corresponds to the delivery pump / valve; b. Color set C = {C1, C2, C3}, corresponding to 3 water quality categories (high salinity: TDS ≥ 10000; medium salinity: TDS 1000-10000; low salinity: TDS ≤ 1000). c. The token function M:P→Bag(C×R) represents the amount of wastewater of different water qualities in the reservoir. For example, M(Psource,1)={C1,80m3 / h} means that water source 1 has 80m³ / h of high-salt wastewater.

[0050] d. Transition triggering conditions (water quality matching + flow rate satisfaction): e. For water point Pwater point,j, its water quality requirement is Creq,j, and if the token color Ci of wastewater source Psource,i is... If Creq,j (water quality matches) and M(Psource,i)≥Qreq,j (flow rate satisfies), then the transition Ti,j (starting the corresponding delivery pump) is triggered.

[0051] ② Formula for calculating water quality matching degree:

[0052] in, / For the conductivity range of wastewater source i, / The conductivity of water point j is within the allowable range. When γi,j≥0.8, it is considered a match.

[0053] ③ Scheduling objective function (minimize water transfer energy consumption + maximize reuse rate): Where Qi,j is the water transport volume from water source i to water point j, Li,j is the water transport distance, λ is the energy consumption coefficient per unit distance, β is the reuse rate weight (taken as 1.5), and ∑Qsource,i is the total wastewater volume.

[0054] (4) Digital twin visualization and early warning layer: Based on the 3D engine, a digital twin model is constructed and mapped 1:1 to the physical water system. The flow direction, water quality parameters and equipment operation status of the whole plant are visualized in real time. At the same time, a fault diagnosis model based on lightweight neural network or statistical learning algorithm is integrated to provide early warning of the risks of pipeline leakage, heat exchanger scaling and water pump efficiency decline, and the early warning accuracy rate is ≥90%. Furthermore, the digital twin visualization and early warning layer is based on the Unity3D engine to build a digital twin model that maps to the physical water system in a 1:1 manner. It integrates functions such as water flow animation simulation, real-time parameter annotation, and fault early warning pop-ups to visualize the dynamic changes of water flow, water quality, and equipment status throughout the plant in real time. It also integrates a fault diagnosis model based on a lightweight neural network to provide early warnings of risks such as pipeline leaks and equipment performance degradation.

[0055] The fault diagnosis model of the digital twin visualization and early warning layer supports online self-learning and continuously optimizes the early warning threshold by accumulating operational data.

[0056] (5) Closed-loop execution layer: Through standardized communication interface, the control commands of the intelligent decision core layer (such as water supply valve opening, sewage valve opening, fan frequency, pump start and stop, wastewater transport path) are sent to the plant-level DCS or sub-control system PLC to drive the action of actuators such as water supply valve, sewage valve, cooling tower fan, and wastewater transport pump, and the execution results are fed back to the data center layer to form a closed-loop control of the whole process of "perception-decision-execution-feedback".

[0057] For example, the closed-loop execution layer automatically controls the water supply valve, drain valve, cooling tower fan frequency converter, wastewater transfer pump, and chemical metering pump of the circulating water system through a plant-level distributed control system or programmable logic controller.

[0058] Optionally, the closed-loop execution layer has a manual / automatic mode switching function, which can be switched to manual control mode under extreme operating conditions.

[0059] In addition, the intelligent control system for dynamic water balance in coal-fired power plants in this application also includes a security protection module, which uses firewall, encrypted transmission, and hierarchical access control technology to ensure the secure transmission of data and control commands.

[0060] The following description, with reference to the accompanying drawings, outlines a water balance control method, apparatus, electronic device, and storage medium for power plants according to embodiments of this application. The water balance control method for power plants is applied to the aforementioned multidimensionally coupled dynamic water balance intelligent control system for coal-fired power plants.

[0061] Specifically, Figure 1 This is a schematic flowchart of a water balance control method for a power plant provided in an embodiment of this application.

[0062] like Figure 1 As shown, the water balance control method of this power plant includes the following steps: In step S101, the current operating status of the power plant's units, the current wastewater quality and quantity information, and the multi-source operating data of the water system are obtained.

[0063] Furthermore, in some embodiments, before inputting the multi-source operational data into a preset time series prediction model, the method further includes: cleaning the multi-source operational data to obtain cleaned multi-source operational data; performing outlier removal processing on the cleaned multi-source operational data to obtain processed multi-source operational data; performing standardization transformation processing on the processed multi-source operational data; and storing the standardized operational data in a preset time series database.

[0064] In some embodiments, the multi-source operating data includes at least one of flow rate, pressure, temperature, pH value, conductivity, turbidity, hardness, and chloride ion content.

[0065] Specifically, in this embodiment, multi-source operational data is collected through a multi-source data perception layer, including flow rate, pressure, temperature, and water quality data of each node in the plant's water system. The data is then transmitted to the edge computing and data hub layer for preprocessing, including data cleaning, outlier removal, standardization conversion, and storage in a time-series database.

[0066] Optionally, water quality data preprocessing also includes normalization of water quality indicators, converting water quality parameters of different dimensions to the [0, 1] interval.

[0067] In step S102, multi-source operating data is input into a preset time-series prediction model, and water volume prediction results are output through the preset time-series prediction model. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and water volume prediction results.

[0068] The water quantity forecast results include at least one of the following: water demand forecast, water production forecast, and reclaimable water quantity forecast.

[0069] Specifically, this application embodiment uses a time-series prediction model to dynamically predict the plant's overall water volume for future periods based on preprocessed data. Based on the water volume prediction results and the current operating status, a multi-objective optimization algorithm is used to solve for the optimal control strategy that satisfies multiple constraints. The training data for the dynamic prediction model includes at least one year of historical operating data from the power plant. The multi-objective particle swarm optimization algorithm is set with a population size of 50-100 and 30-50 iterations.

[0070] The optimization variables of the multi-objective optimization algorithm preset in the embodiments of this application are: X=[K,V f a1, a2, ... a m ], where K is the circulating water concentration ratio, V f For the cooling tower fan frequency, a1~a m Set the opening degree of each water supply / drainage valve.

[0071] Furthermore, the objective function for optimizing the variables is obtained.

[0072] (1) Water consumption per unit of power generation f1=Q fresh / P, where Q is the fresh water intake (m³ / h) and P is the unit's power generation (MW·h). (2) Auxiliary power consumption of water intake, drainage, and water use systems. , where P i Let ρ be the power of the i-th pump (kW), and ρ be the density of water (kg / m³). 3 g is the acceleration due to gravity (m / s²). 2 ), Q i Let the flow rate of the i-th pump be (m³ / s). 3 / s), Hi is the head (m) of the i-th pump, η i For the efficiency of the i-th pump, η motor For motor efficiency; (3) Pollution load of discharged wastewater Q out,j The flow rate of wastewater discharged from the j-th channel is (m³ / h). The concentration of the nth pollutant in the jth wastewater (mg / L, such as COD, chloride ions, etc.).

[0073] Furthermore, the dynamic weight fusion objective is: F(X) = W1f1 + W2f2 + W3f3, where W1 + W2 + W3 = 1, and the dynamic weight coefficients take the following values: a. Rated load condition (load ≥ 75% of rated load): W1 = 0.5, W2 = 0.3, W3 = 0.2; Normal regulation zone (50% of rated load ≤ load < 75% of rated load): W1 = 0.4, W2 = 0.3, W3 = 0.3; Deep peak shaving condition (load ≤ 50% of rated load): W1 = 0.2, W2 = 0.5, W3 = 0.3.

[0074] b. Constraints: Hard constraints: 2.5≤K≤5.0 (circulating water concentration ratio), 7.5≤pH≤9.5 (circulating water pH), 0≤αi≤100% (valve opening); Soft constraints: fluctuation range of key water quality parameters under deep peak shaving conditions ≤±5%.

[0075] Furthermore, PSO particle updates: a. Particle velocity update: ; b. Particle position update: ; Where ω is the inertia weight (linearly decreasing from 0.4 to 0.9), c1 and c2 are learning factors (both set to 2), and r1 and r2 are random numbers in the range [0,1]. Let be the optimal position for the i-th particle. This is the globally optimal position for the population.

[0076] Furthermore, the algorithm performance is constrained as follows: the optimization cycle is ≤10 seconds, and the optimization cycle is ≤5 seconds under deep peak shaving conditions. The output includes a set of control parameters, including the target value of the circulating water concentration ratio, the reference value of the cooling tower fan frequency, and the opening command of the key water supply valve.

[0077] In step S103, the current wastewater quality and quantity information is input into the preset wastewater scheduling model, and wastewater utilization planning is carried out through the preset wastewater scheduling model to generate the target wastewater utilization strategy.

[0078] Furthermore, in some embodiments, the preset wastewater scheduling model has the following objective function: ; in, Q i,j Let i be the amount of water transported from water source i to water point j. L i,j Let λ be the water transport distance from water source i to water point j, and λ be the energy consumption coefficient per unit distance. β For the reuse rate weight, ∑ Q source,i This represents the total wastewater volume.

[0079] Specifically, this application embodiment calculates the optimal wastewater cascade utilization scheme in real time using a dynamic scheduling model based on the current wastewater quality and quantity information. When planning wastewater utilization through a preset wastewater scheduling model, the principle of "high-quality wastewater is used high, low-quality wastewater is used low" is followed, prioritizing the reuse of wastewater with better water quality in the circulating water makeup system.

[0080] In detail, the embodiments of this application are based on a wastewater scheduling model constructed using colored Petri nets. This model abstracts the complex wastewater system into a computable and deductive dynamic network. First, water quality parameters and flow data of various wastewater sources (such as reverse osmosis concentrate, chemical backwash water, desulfurization wastewater, etc.) throughout the plant are collected in real time to form current wastewater quality and quantity information. This information is converted into "tokens" in the model, where the color attribute of the tokens is used to distinguish different water quality categories. For example, purple represents high-salinity wastewater, blue represents medium-salinity wastewater, and green represents low-salinity wastewater. The number of tokens represents the instantaneous flow rate of the corresponding wastewater. At the same time, the preset "reservoir" nodes in the model correspond to various wastewater sources, potential water use points, and connecting pipelines, while the "transition" nodes correspond to the start-up, shutdown, or adjustment actions of pumps, valves, and other conveying actuators.

[0081] The wastewater scheduling model automatically plans dynamic paths for wastewater utilization based on preset scheduling rules and optimization objectives. The planning process follows the fundamental principle of "high-quality, high-use; low-quality, low-use," with water quality matching and flow balance calculations as its core. The model compares the water quality characteristics of each wastewater source with the water quality requirements of each water usage point (such as circulating water makeup systems, coal flushing, and dry slag humidification). For example, when a water usage point allows the use of wastewater with medium or lower salinity, the model dynamically calculates one or more optimal transport paths from all available water sources that meet this water quality condition, based on factors such as real-time flow rate, transport distance, and energy consumption. This calculation process is simulated by triggering corresponding "transitions" in the Petri net. The triggering condition for these transitions is strictly set to "water quality matching and flow rate satisfaction," ensuring the feasibility of the scheduling scheme. The objective function of the scheduling optimization aims to minimize the total energy consumption of water transport while maximizing the wastewater reuse rate, thereby achieving the best balance between energy and water conservation.

[0082] Ultimately, the wastewater scheduling model outputs a specific target wastewater utilization strategy. This strategy is a set of executable instructions that clearly specifies which wastewater source, which point of use, how much flow rate should be delivered, and which pumps or valves should be started or stopped to achieve this delivery in the next control cycle. This enables dynamic and optimal allocation of wastewater resources, effectively improves the plant's water resource reuse rate, and reduces the amount of fresh water taken out and the amount of wastewater discharged.

[0083] In step S104, a target control command is generated based on the target control strategy and the target wastewater utilization strategy, and the target control command is sent to the actuator to control the actuator to perform corresponding adjustment actions based on the target control command.

[0084] Specifically, the optimal control strategy and wastewater utilization scheme are integrated to generate control commands and issue them to the actuators, completing closed-loop control. In detail, after completing the optimization calculations, the intelligent decision-making core layer integrates, encodes, and formats the core parameters in the target control strategy (such as the target value of the circulating water concentration ratio, the target frequency of the cooling tower fan, and the target opening degree of the key water supply valve and drain valve) with the specific path instructions in the target wastewater utilization strategy (such as the specified wastewater transfer pump number, valve opening degree, and target flow rate), forming target control commands. These commands are then transmitted securely and reliably through the plant-level data bus or a dedicated control network.

[0085] The command issuance process is completed through a closed-loop execution layer, sending the command to the corresponding actuator. The actuators encompass various key equipment in the water system, primarily including regulating valves (such as water supply valves and drain valves) and power equipment (such as cooling tower fan frequency converters, various wastewater transfer pumps, and chemical metering pumps). Upon receiving the target control command, the actuator immediately drives its mechanical or electrical components to perform the corresponding adjustment actions. For example, a water supply regulating valve will adjust its opening to the percentage specified in the command; a cooling tower fan will adjust its operating frequency to the target Hertz value; and a group of wastewater transfer pumps will start, stop, or adjust their speed according to the command requirements, transporting wastewater of specified quality and flow rate along the planned path to the target water usage point.

[0086] Furthermore, in some embodiments, after acquiring the current unit operating status of the power plant, the current wastewater quality and quantity information, and the multi-source operating data of the water system, the method further includes: monitoring the operating status of the water system; generating early warning information when the operating status of the water system meets preset risk conditions, and issuing risk warnings based on a preset digital twin platform; wherein the early warning information includes at least one of fault location information and risk handling strategies.

[0087] Specifically, this application embodiment visualizes the entire control process in a digital twin platform and runs a fault early warning model. Through a lightweight neural network fault diagnosis model integrated in the twin platform, it continuously analyzes massive amounts of information from the multi-source data perception layer, calculates key health indicators in real time such as pipeline flow-pressure matching degree, heat exchanger terminal difference and efficiency, and water pump performance curve deviation, and compares them with the normal operation mode thresholds obtained through training and learning from historical big data.

[0088] When real-time calculations and model diagnostics identify that the operating status of the water system meets preset risk conditions, an early warning mechanism will be automatically triggered. For example, the preset pipeline leakage risk condition simultaneously meets multiple related criteria such as an abnormal increase in the flow meter reading of a certain branch, an abnormal decrease in the corresponding pressure sensor reading, and abnormal water level or humidity monitoring in that area; the heat exchanger scaling risk condition can be a combination of trends such as "continuously decreasing heat exchange temperature difference", "continuously increasing pressure loss under constant flow rate", and "thermal efficiency index continuously below the threshold". When such composite conditions are met, the risk is determined to be established.

[0089] Furthermore, structured early warning information is dynamically generated. This information not only includes traditional alarm levels, times, and locations, but also integrates diagnostic results, including at least one or more of the following: fault location information and risk handling strategies. Fault location information can be accurate to the specific equipment, pipe section, or even bolt location in the 3D twin model, and is highlighted. Risk handling strategies are driven by a knowledge base and case library, providing preliminary handling suggestions, such as "Suspected internal leakage in the outlet valve of #3 circulating water pump; it is recommended to prioritize checking valve position feedback and arranging isolation verification" or "#1 cooling tower packing scaling trend warning; it is recommended to increase the sewage discharge rate and check the chemical concentration." Early warning information can be delivered through the digital twin platform in various ways, such as pop-ups, audio-visual presentations, 3D model highlighting and flashing, and trend curve highlighting, and is simultaneously pushed to the mobile terminals of operators and the duty system, ensuring that risks are detected in a timely manner, accurately located, and effectively handled, thereby realizing the transformation from post-fault maintenance to pre-risk warning intelligent operation and maintenance mode.

[0090] Furthermore, in this embodiment of the application, based on the deviation between the actual feedback data of the system after closed-loop execution and the prediction optimization results, some parameters of the time-series prediction model and / or multi-objective optimization model are adaptively corrected online.

[0091] To help those skilled in the art to better understand the water balance control method for power plants described in this application, the following explanation is provided with examples.

[0092] The water balance control method of this power plant includes the following steps: S1: Real-time acquisition and preprocessing of multi-source data. The multi-source data sensing layer collects flow, pressure, temperature and water quality data of each node in the plant's water system, transmits them to the edge computing and data hub layer to complete data cleaning, outlier removal and standardization transformation, and stores them in the time series database. S2: Dynamic prediction of water quantity and quality. It calls the bidirectional long short-term memory network-attention composite model of the dynamic prediction module to predict the water demand, water production and reclaimable water volume of each subsystem in the next 30-60 minutes based on the pre-processed data. S3: Multi-objective collaborative optimization solution, calling the MOPSO (Multi-objective Particle Swarm Optimization) algorithm of the multi-objective optimization module, taking the prediction results of step S2 as input, and combining the current unit operating status and constraints to solve and generate the optimal control strategy set; S4: Wastewater cascade scheduling path planning, calling the colored Petri net model of the wastewater dynamic scheduling module, and calculating the optimal wastewater reuse path and flow rate based on real-time wastewater source and water quality and quantity data of water use points; S5: Control command issuance and closed-loop execution, integrating the optimal control strategy of step S3 and the wastewater scheduling scheme of step S4, generating standardized control commands and issuing them to the actuators, while collecting execution results and feeding them back to the data hub layer; S6: Digital twin visualization and risk warning. The system operation status is displayed in real time on the digital twin platform, the fault diagnosis model is called to identify and warn of risks, and fault location and handling suggestions are pushed. S7: Online adaptive model update. Based on the deviation between the execution feedback data of step S5 and the prediction results of step S2 and the optimization results of step S3, the parameters of the dynamic prediction model and the multi-objective optimization model are corrected online. The online adaptive model update adopts an incremental learning algorithm, which does not require retraining all historical data. S8: Robust control under deep peak shaving conditions. Under deep peak shaving conditions, when the unit load is detected to be lower than 40% of the rated load, the system automatically switches to enhanced robust control mode, increases the optimization calculation frequency and relaxes some non-core constraints to ensure that the fluctuation range of key parameters of the water system is controlled within ±5%. The enhanced robust control mode increases the optimization calculation frequency to once every 5 seconds. S9: Operational data archiving and analysis. Regularly archive system operation data, control strategies, and benefit data to provide data support for subsequent technical optimization and benefit evaluation.

[0093] Therefore, this application provides a system and method capable of dynamic perception, real-time prediction, multi-objective optimization, and intelligent closed-loop control of the entire power plant's water system. Driven by real-time data, it utilizes an LSTM model for dynamic water volume prediction, employs the MOPSO algorithm for multi-objective collaborative optimization of water conservation, energy saving, and environmental protection, and achieves dynamic intelligent scheduling of wastewater based on a colored Petri net. Finally, it uses a digital twin platform for visualized monitoring and command issuance to form closed-loop control. This solves the problems of static, isolated, and extensive water management in traditional power plants, achieving real-time, precise, collaborative, and intelligent control of the entire plant's water system. This significantly improves water resource utilization efficiency and operational economy, promoting digital transformation and intelligent upgrading in the industrial sector.

[0094] According to the water balance control method for power plants proposed in this application, the multi-source operation data of the current power plant water system is input into a preset time-series prediction model, which outputs water volume prediction results. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and water volume prediction results. The current wastewater quality and quantity information is input into a preset wastewater scheduling model for wastewater utilization planning, generating a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy and sent to the actuators to control them to perform corresponding adjustment actions based on the target control instructions. This overcomes the static, isolated, and extensive defects of existing coal-fired power plant water system management, significantly reducing water intake per unit of power generation, improving the comprehensive utilization efficiency of water resources, and adapting to complex operating conditions such as deep peak shaving.

[0095] Next, the water balance control device for a power plant according to an embodiment of this application is described with reference to the accompanying drawings.

[0096] Figure 2 This is a block diagram of a water balance control device for a power plant according to an embodiment of this application.

[0097] like Figure 2 As shown, the water balance control device 10 of the power plant includes: an acquisition module 100, a processing module 200, a wastewater scheduling module 300, and a control module 400.

[0098] The system comprises the following modules: an acquisition module 100, which acquires the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system; a processing module 200, which inputs the multi-source operating data into a preset time-series prediction model, outputs water quantity prediction results through the preset time-series prediction model, and, based on a preset multi-objective optimization algorithm, obtains a target control strategy that meets preset constraints according to the current unit operating status and water quantity prediction results; a wastewater scheduling module 300, which inputs the current wastewater quality and quantity information into a preset wastewater scheduling model, performs wastewater utilization planning through the preset wastewater scheduling model, and generates a target wastewater utilization strategy; and a control module 400, which generates target control instructions based on the target control strategy and the target wastewater utilization strategy, and sends the target control instructions to the actuators to control the actuators to perform corresponding adjustment actions based on the target control instructions.

[0099] Furthermore, in some embodiments, the objective function of the preset multi-objective optimization algorithm is: F(X) = W 1 f 1 +W 2 f 2 +W 3 f 3; f1=Qfresh / P ; ; ; Where F(X) is the objective function, f 1 represents the water intake per unit of power generation. f 2 represents auxiliary power consumption. f 3 represents the pollution load of discharged wastewater. Q fresh This refers to the amount of fresh water taken. P For the power generation of the unit, Let be the power of the i-th pump. For the density of water, It is the acceleration due to gravity. Let i be the flow rate of the i-th pump. Let i be the head of the i-th pump. Let the efficiency of the i-th pump be... For motor efficiency, For the j-th discharge wastewater flow rate, Let n be the concentration of the nth pollutant in the j-th wastewater. W 1. W 2. W 3 are all weighting coefficients.

[0100] Furthermore, in some embodiments, the preset wastewater scheduling model has the following objective function: ; in, Q i,j Let i be the amount of water transported from water source i to water point j. L i,j Let λ be the water transport distance from water source i to water point j, and λ be the energy consumption coefficient per unit distance. β For the reuse rate weight, ∑ Q source,i This represents the total wastewater volume.

[0101] Furthermore, in some embodiments, after acquiring the current unit operating status of the power plant, the current wastewater quality and quantity information, and the multi-source operating data of the water system, the acquisition module 100 is also used to: monitor the operating status of the water system; generate early warning information when the operating status of the water system meets preset risk conditions, and provide risk early warning based on a preset digital twin platform; wherein the early warning information includes at least one of fault location information and risk handling strategies.

[0102] Furthermore, in some embodiments, before inputting the multi-source operational data into the preset time series prediction model, the processing module 200 is further configured to: clean the multi-source operational data to obtain cleaned multi-source operational data; perform outlier removal processing on the cleaned multi-source operational data to obtain processed multi-source operational data; perform standardization transformation processing on the processed multi-source operational data, and store the standardized operational data in the preset time series database.

[0103] Furthermore, in some embodiments, the multi-source operating data includes at least one of flow rate, pressure, temperature, pH value, conductivity, turbidity, hardness, and chloride ion content, and the water volume prediction results include at least one of water demand prediction, water production prediction, and reclaimable water volume prediction.

[0104] It should be noted that the foregoing explanation of the water balance control method embodiment for power plants also applies to the water balance control device of the power plant in this embodiment, and will not be repeated here.

[0105] According to the water balance control device for power plants proposed in this application, the multi-source operating data of the current power plant water system is input into a preset time-series prediction model, which outputs water volume prediction results. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and water volume prediction results. The current wastewater quality and quantity information is input into a preset wastewater scheduling model for wastewater utilization planning, generating a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy and sent to the actuators to control the actuators to perform corresponding adjustment actions based on the target control instructions. This overcomes the static, isolated, and extensive defects of existing coal-fired power plant water system management, significantly reducing water intake per unit of power generation, improving the comprehensive utilization efficiency of water resources, and adapting to complex operating conditions such as deep peak shaving.

[0106] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 301, the processor 302, and the computer program stored on the memory 301 and capable of running on the processor 302.

[0107] When the processor 302 executes the program, it implements the power plant water balance control method provided in the above embodiments.

[0108] Furthermore, electronic devices also include: Communication interface 303 is used for communication between memory 301 and processor 302.

[0109] The memory 301 is used to store computer programs that can run on the processor 302.

[0110] The memory 301 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0111] If the memory 301, processor 302, and communication interface 303 are implemented independently, then the communication interface 303, memory 301, and processor 302 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0112] Optionally, in a specific implementation, if the memory 301, processor 302, and communication interface 303 are integrated on a single chip, then the memory 301, processor 302, and communication interface 303 can communicate with each other through an internal interface.

[0113] Processor 302 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0114] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described water balance control method for a power plant.

[0115] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0116] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0117] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0118] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0119] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0120] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0121] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0122] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for water balance control in a power plant, characterized in that, Includes the following steps: Obtain the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system; The multi-source operating data is input into a preset time-series prediction model, and the water volume prediction result is output through the preset time-series prediction model. Based on a preset multi-objective optimization algorithm, a target control strategy that meets preset constraints is obtained according to the current unit operating status and the water volume prediction result. The current wastewater quality and quantity information is input into a preset wastewater scheduling model, and wastewater utilization planning is carried out through the preset wastewater scheduling model to generate a target wastewater utilization strategy. Target control instructions are generated based on the target control strategy and the target wastewater utilization strategy, and then sent to the actuator to control the actuator to perform corresponding adjustment actions based on the target control instructions.

2. The method according to claim 1, characterized in that, The objective function of the preset multi-objective optimization algorithm is: F(X)= W 1 f 1 +W 2 f 2 +W 3 f 3; f1=Q fresh / P ； ； ； Where F(X) is the objective function, f 1 represents the water intake per unit of power generation. f 2 represents auxiliary power consumption of the water system. f 3 represents the pollution load of discharged wastewater. Q fresh This refers to the amount of fresh water taken. P For the power generation of the unit, Let be the power of the i-th pump. For the density of water, It is the acceleration due to gravity. Let i be the flow rate of the i-th pump. Let i be the head of the i-th pump. Let the efficiency of the i-th pump be... For motor efficiency, For the j-th discharge wastewater flow rate, Let n be the concentration of the nth pollutant in the j-th wastewater. W 1. W 2. W 3 are all weighting coefficients.

3. The method according to claim 1, characterized in that, The objective function of the preset wastewater scheduling model is: ； in, Q i,j Let i be the amount of water transported from water source i to water point j. L i,j Let λ be the water transport distance from water source i to water point j, and λ be the energy consumption coefficient per unit distance. β For the reuse rate weight, ∑ Q source,i This represents the total wastewater volume.

4. The method according to claim 1, characterized in that, After acquiring the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system, it also includes: Monitor the operating status of the water system; When the operating status of the water system meets the preset risk conditions, an early warning reminder is generated, and a risk warning reminder is issued based on the preset digital twin platform; The early warning information includes at least one of fault location information and risk handling strategies.

5. The method according to claim 1, characterized in that, Before inputting the multi-source operational data into the preset time-series prediction model, the method further includes: The multi-source operational data is cleaned to obtain cleaned multi-source operational data; The cleaned multi-source operating data is subjected to outlier removal processing to obtain processed multi-source operating data; The processed multi-source operational data is standardized and converted, and the standardized operational data is stored in a preset time-series database.

6. The method according to claim 1, characterized in that, The multi-source operating data includes at least one of flow rate, pressure, temperature, pH value, conductivity, turbidity, hardness, and chloride ion content, and the water volume prediction results include at least one of water demand prediction, water production prediction, and reclaimable water volume prediction.

7. A water balance control device for a power plant, characterized in that, include: The acquisition module is used to acquire the current operating status of the power plant's units, current wastewater quality and quantity information, and multi-source operating data of the water system. The processing module is used to input the multi-source operating data into a preset time-series prediction model, output water volume prediction results through the preset time-series prediction model, and obtain a target control strategy that meets preset constraints based on a preset multi-objective optimization algorithm, according to the current unit operating status and the water volume prediction results. The wastewater scheduling module is used to input the current wastewater quality and quantity information into a preset wastewater scheduling model, and to perform wastewater utilization planning through the preset wastewater scheduling model to generate a target wastewater utilization strategy. The control module is used to generate target control instructions based on the target control strategy and the target wastewater utilization strategy, and to send the target control instructions to the actuator to control the actuator to perform corresponding adjustment actions based on the target control instructions.

8. The apparatus according to claim 7, characterized in that, The objective function of the preset multi-objective optimization algorithm is: F(X)= W 1 f 1 +W 2 f 2 +W 3 f 3; f1=Q fresh / P ； ； ； Where F(X) is the objective function, f 1 represents the water intake per unit of power generation. f 2 represents auxiliary power consumption. f 3 represents the pollution load of discharged wastewater. Q fresh This refers to the amount of fresh water taken. P For the power generation of the unit, Let be the power of the i-th pump. For the density of water, It is the acceleration due to gravity. Let i be the flow rate of the i-th pump. Let i be the head of the i-th pump. Let the efficiency of the i-th pump be... For motor efficiency, For the j-th discharge wastewater flow rate, Let n be the concentration of the nth pollutant in the j-th wastewater. W 1. W 2. W 3 are all weighting coefficients.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, the processor executing the computer program to implement the water balance control method for a power plant as described in any one of claims 1-6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor to implement the water balance control method for a power plant as described in any one of claims 1-6.

Images