CN-121998235-A - Intelligent management and auxiliary decision-making method for water conservancy information

CN121998235ACN 121998235 ACN121998235 ACN 121998235ACN-121998235-A

Abstract

The application relates to an intelligent management and auxiliary decision-making method for water conservancy information, which comprises the following steps of collecting dynamic information of water resources in water conservancy equipment and node terminals, synchronously calibrating the dynamic information by utilizing a multi-source data fusion technology, processing the dynamic information of the water resources, extracting key characteristic data based on a distributed processing architecture, analyzing the key characteristic data by utilizing edge computing equipment in combination with a dynamic optimization model to generate a prediction result of the state of the water resources, generating a local management decision-making strategy by utilizing a self-adaptive algorithm in combination with a set rule base according to the characteristic index collected in real time and the stored analysis result, verifying and adjusting the local management decision-making strategy by utilizing a multi-target optimization algorithm in a centralized management system, and transmitting the optimized global management decision-making strategy to an executing terminal device.

Inventors

XU JUNYANG
SHOU WEIWEI
XU WEIKANG
GU CHENHUI
JIANG QIN
YU QUAN
CHEN ZEYANG
QIAN YAJUN
CAI PENG

Assignees

杭州华辰电力控制工程有限公司

Dates

Publication Date: 20260508
Application Date: 20251219

Claims (10)

1. The intelligent management and decision-making assisting method for the water conservancy information is characterized by comprising the following steps of: S1, constructing a distributed sensing network, collecting dynamic information of water resources in water conservancy equipment and node terminals, and synchronously calibrating the sensing information by utilizing a multi-source data fusion technology; s2, configuring edge computing equipment at a water conservancy node terminal, transmitting the acquired water resource dynamic information to the edge computing equipment, performing data cleaning, outlier rejection and format standardization on the water resource dynamic information, and extracting key feature data based on a distributed processing architecture; S3, analyzing the key feature data by utilizing the edge computing equipment and combining a dynamic optimization model to generate a prediction result of the water resource state, extracting feature indexes related to decision and storing the prediction result in a local storage unit; S4, generating a local management decision strategy by combining a self-adaptive algorithm with a set rule base according to the characteristic indexes acquired in real time and the stored prediction results, and transmitting the locally generated local management decision strategy to a centralized management system through a communication network; s5, verifying and adjusting the local management decision strategy by utilizing a multi-objective optimization algorithm in the centralized management system to generate a global management decision strategy meeting optimization requirements; And S6, transmitting the optimized global management decision strategy to the execution end equipment, controlling related water conservancy equipment to implement management decisions, monitoring data changes in the implementation process in real time through a sensing network, and recording an execution result and environmental response.
2. The method for intelligent management and decision-making assistance of water conservancy information according to claim 1, wherein the step S1 specifically comprises the following steps: S11, distributing intelligent sensing nodes supporting distributed cooperation in water conservancy equipment and node terminals to acquire dynamic information of water resources, wherein the intelligent sensing nodes comprise a water level sensor, a flow rate sensor, a water quality sensor and an air image sensor; S12, when the dynamic information of the water resource is acquired, calculating and monitoring the change rate of the dynamic parameters of the water resource in the environment in real time According to the calculated change rate Combining the adjustment coefficients And minimum sampling period Dynamically determining sampling periods : ; In the formula, As a function of the dynamic parameters, As time variable, dynamic parameters include water level, flow rate and water quality index, Representing a minimum sampling period; Setting a threshold range for the adjustment process of the sampling period Limiting the variation amplitude of two successive sampling periods: ; in the formula, And Respectively represent the first Subsampling period and the th The sub-sampling period is taken to be a sub-sampling period, Is a preset variation range threshold value; S13, based on a space-time coordination mechanism of the sensing nodes, performing time calibration on the acquired dynamic information of the water resource by adopting a time synchronization protocol, and unifying geographic reference coordinates of measurement data through a spatial position correction model among the nodes; s14, calculating the dynamic weight of each observation data aiming at the multi-source observation data in the water resource dynamic information And according to dynamic weights And generating a collaborative processing result after fusion by the observation data.
3. The method for intelligent management and decision-making assistance of water conservancy information according to claim 2, wherein the step S14 specifically comprises the following steps: multi-source observation data in dynamic information of water resources According to the environment priori state Establishing a conditional probability model Expressed in a given prior state The credibility of each sensor data is calculated, and the dynamic weight of each observed data is calculated by using a Bayesian inference method : ; In the formula, Is the first The dynamic weights of the individual observations are such that, For the confidence probability of the observed data in the prior-verification state, A total number of sensors involved in the cooperative processing; By a dynamic updating mechanism, based on the accuracy of sensor historical data, the current environmental noise level and the working state of the sensor, real-time adjustment is performed Is the value of (1): ; in the formula, Represent the first The reliability score of the individual sensors, Representing a value prior to a state update; According to the calculated dynamic weight Weighting and fusing all the observed data to generate a fused cooperative processing result: ; in the formula, As the observation data after the fusion, Is the first Observations of the individual sensors.
4. The method for intelligent management and decision-making assistance of water conservancy information according to any one of claims 1 to 3, wherein the step S2 specifically comprises the following steps: S21, transmitting the dynamic information of the water resource acquired through the distributed sensing network to the edge computing equipment configured at the water conservancy node terminal in the form of a data frame, wherein the data frame comprises a time stamp, a sensor identifier and a dynamic parameter value; S22, in the edge computing equipment, data cleaning is carried out on the received dynamic information, and incomplete data frames and abnormal data exceeding a threshold range are removed; S23, carrying out format standardization processing on the cleaned data frame, and converting dynamic information into structural data containing multi-dimensional attribute fields based on a predefined unified data model, wherein the multi-dimensional attribute fields comprise a time stamp, space coordinates, dynamic parameter values and a sensor identifier; S24, based on a distributed processing architecture, dividing the structured data after format standardization into a plurality of data partitions, wherein a data partition rule is set according to a space range and a time period, and a partition index is set The method comprises the following steps: ; in the formula, For the partition index to be used, The spatial coordinates of the data are represented, In order to be a time stamp, 、 And Representing the partition intervals of space and time respectively, And Respectively representing the number of the space partitions; S25, in each data partition, based on dynamic parameter set Extracting key feature data by adopting a high-order difference method: ; in the formula, In order to dynamically change the rate of change, For the differential order number, In order to sample the time interval of the time, Is the number of combinations.
5. The method for intelligent management and decision-making assistance of water conservancy information according to any one of claims 1 to 3, wherein the step S3 comprises the following steps: S31, loading a dynamic optimization model in the edge computing equipment, and analyzing extracted key feature data, wherein the key feature data comprises a dynamic parameter set and a corresponding time stamp; s32, when a water resource state prediction equation is constructed based on the dynamic optimization model, adopting a time sequence prediction algorithm to perform dynamic parameter set Modeling in which time series data is fitted using an autoregressive integral-moving average model, predicted values The method comprises the following steps: ; in the formula, In order to be able to predict the value, And The parameters of the autoregressive and moving average models respectively, And For the order of the model, the number of the model, In order to predict the error of the signal, Optimizing model parameters for bias terms by fitting to historical time series data 、 And In the prediction process, the time series data is updated in real time by introducing a sliding window technology, and only the window length is reserved each time Is the latest data of (a) ; S33, predicting result Comparing with the real-time collected data, and adjusting the parameters of the dynamic optimization model by using an error correction method; s34, extracting feature indexes related to decision in the prediction result, and based on the prediction value set And dynamically changing parameters to construct a feature extraction function set Each feature extraction function Calculating specific indexes according to different decision requirements, performing preliminary screening on all calculated characteristic indexes, filtering invalid and redundant characteristics according to constraint conditions and priorities set in a rule base R, and finally screening to obtain a key characteristic index set Wherein each key index Meets the rule And storing according to the order of decision priority thereof; And S35, storing the prediction results of the analyzed and screened characteristic index set and the water resource state to a local storage unit of the edge computing equipment.
6. The method for intelligent management and decision-making assistance of water conservancy information according to claim 5, wherein the step S33 specifically comprises the following steps: For prediction errors Constructing a gradient descent-based parameter optimization process: ; model parameters by calculating error functions 、 And Adopts iteration update rule to optimize parameters: ; ; ; in the formula, In order for the rate of learning to be high, 、、 Respectively the partial derivatives of the prediction errors to model parameters; An adaptive learning rate adjustment mechanism is introduced to dynamically adjust the learning rate according to the current error change rate Efficiency and stability of error correction are optimized: ; in the formula, In order to adjust the coefficient of the coefficient, Representing the amount of change in error in two iterations, A symbol representing a change in error.
7. The method for intelligent management and decision-making assistance of water conservancy information according to any one of claims 1 to 3 or 6, wherein the step S4 specifically comprises the following steps: s41, feature index set based on real-time acquisition And stored analysis result sets Initializing input parameters of an adaptive algorithm, including an index weight vector Optimizing an objective function ; S42, combining rule base In the method, a local management decision objective function is constructed through a multi-objective optimization model ; S43, adopting genetic algorithm to make local management decision objective function Optimizing; s44, optimizing the genetic algorithm to obtain an optimal local management decision parameter set Mapping to specific management decision policies The mapping process is based on a rule base Policy generation rules in (1) to generate each optimal parameter value Conversion to specific management actions or control instructions, the mapped management decision strategy set is expressed as Wherein each policy is associated with a corresponding parameter in the objective function Associating; S45, dynamically adjusting rule base based on real-time environment feedback data Constraint conditions in (a) And index weight By feeding back data in real time Calculating the execution deviation of the current management strategy Introducing the deviation into a rule base, and dynamically updating constraint condition functions And the weight of the objective function According to the rate of change of the feedback data Real-time adjustment of learning rate used in optimization process Constraint penalty coefficient ; S46, local management decision strategy to be generated And storing the policies in a local storage unit of the edge computing device and transmitting the policies to the centralized management system through a communication network.
8. The intelligent management and decision-making method for water conservancy information according to claim 7, wherein the step S42 specifically comprises the following steps: Binding rule base In (3) determining a feature index set And analyzing the result set Initializing index weights Weighting of The constraint conditions are satisfied: ; Calculating characteristic index And analysis result Is a matching degree function of (2) : ; In the formula, Adjusting the characteristic index and the matching sensitivity of the analysis result for the scale parameter of the matching degree function; According to rule base Constraint conditions set in (3) Defining constraint functions The method comprises the following steps: ; in the formula, As the weight coefficient of the light-emitting diode, Is the first Constraint conditions Measuring the satisfaction degree of the current state on the constraint condition by the deviation function of the current state; To match the degree function Constraint condition function Substituting local management decision objective function : ; In the formula, For constraint punishment coefficients, dynamically adjusting through a rule base, and balancing weights between the matching degree and constraint conditions; updating parameters in an objective function based on real-time feedback, including dynamically adjusting index weights Constraint penalty coefficient To make the objective function Is consistent with the real-time data of the current water resource status.
9. The intelligent management and decision-making method for water conservancy information according to claim 7, wherein the step S43 specifically comprises the following steps: initializing a population Wherein each individual Represents a group of candidate decision parameters, the population size is The individual codes are expressed by real codes, and each coding parameter corresponds to a local management decision objective function Is a variable of the input of (a); based on objective functions For each individual in the population Evaluating the adaptability and the adaptability function Representing the performance of the individual on the optimization objective: ; in the formula, In individuals as objective functions The value of the above-mentioned value, Is the current optimal target value; Selecting population according to fitness value, determining parent individuals of next generation by adopting roulette selection method, and selecting probability The method comprises the following steps: ; in the formula, For individuals Is used for the adaptation value of the (a), Summing all individual fitness values; Performing cross operation on the selected parent individuals, performing random exchange on the codes of the parent individuals by adopting a single-point cross method, and generating next generation candidate solutions and cross probabilities Controlling the incidence of crossover operations; performing mutation operation on the individuals generated after crossing, and generating a new solution space by randomly changing a certain parameter value in individual codes, wherein the mutation probability Controlling the occurrence rate of mutation operation: ; in the formula, As the factor of the amplitude of the variation, To be in the interval A numerical value randomly generated in the memory; Evaluating the fitness of the mutated population, screening out the individual with the highest fitness value as the optimal solution of the current population, and recording the optimal solution as And update the objective function Optimum value of (2) 。
10. The intelligent management and decision-making method for water conservancy information according to claim 7, wherein the step S5 specifically comprises the following steps: s51, receiving a local management decision strategy set and global water resource dynamic data acquired in real time For local strategy With global data Performing consistency verification; S52, combining with global optimization requirements, and constructing a global management decision objective function based on a multi-objective optimization algorithm : ; In the formula, Representing local policies The degree of adaptation under global dynamic data, As a weight of the local policy, Representing global constraints Is a function of the degree of deviation of (2), Punishment coefficients for global constraints; s53, utilizing a multi-objective optimization algorithm to globally manage decision objective functions Optimizing; s54, collecting the optimized global management decision parameters Mapping to specific global management decision strategy Each policy By optimizing parameters Converted into; S55, generating a global management decision strategy And distributing the data to related water conservancy equipment and node terminals through a communication network.

Description

Intelligent management and auxiliary decision-making method for water conservancy information Technical Field The application relates to the technical field of water conservancy resource management, in particular to an intelligent water conservancy information management and auxiliary decision-making method. Background With the rapid development of information technology and intelligent management technology, the informatization level of the water conservancy industry is continuously improved, and the management and decision of water resources gradually progress to automation and intelligent directions. However, the dynamic change of water resources is complex and widely distributed, and the conventional water conservancy information management method still has many challenges in terms of processing massive data, dynamic perception and auxiliary decision support. In the related art, the traditional water conservancy information management and decision support method mainly depends on a centralized data acquisition and manual analysis mode. While these approaches can play a role in certain scenarios, in the face of current water resource management complexity and emergency response needs, conventional techniques have significant shortcomings in several respects: 1. The existing water conservancy information management system mostly adopts a centralized data processing architecture, the period from data acquisition to analysis is long, and the dynamic change of water resources cannot be reflected in real time. This delay makes it difficult for conventional methods to quickly cope with emergency events such as flood early warning and real-time monitoring and decision making of water pollution. 2. The data processing efficiency is low, and mass data are generated by a large number of distributed sensors along with the development of the technology of the Internet of things. However, the conventional method lacks of distributed and edge computing capability, and has huge bandwidth and computing resource consumption for data transmission and centralized processing, and cannot efficiently process multi-source and multi-dimensional large-scale dynamic data. 3. The traditional water conservancy management system depends on static rules or experience decision models, the methods cannot fully utilize real-time dynamic data and historical data to carry out comprehensive analysis, and the method lacks scientific auxiliary decision making capability and is not attractive in multi-objective optimization and regulation and control of complex scenes. 4. The sensing perception is incomplete, namely the existing sensing network is mostly based on a single type of sensor, and the accurate perception and dynamic fusion of the multidimensional water resource information cannot be realized, so that the incompleteness and the accuracy of the decision information are insufficient. Disclosure of Invention The application provides an intelligent management and auxiliary decision-making method for water conservancy information, which has the advantages of strong real-time performance, high data processing efficiency, strong scientificity for decision support and strong environmental adaptability. The intelligent management and auxiliary decision-making method for the water conservancy information provided by the application comprises the following steps: S1, constructing a distributed sensing network, collecting dynamic information of water resources in water conservancy equipment and node terminals, and synchronously calibrating the sensing information by utilizing a multi-source data fusion technology; s2, configuring edge computing equipment at a water conservancy node terminal, transmitting the acquired water resource dynamic information to the edge computing equipment, performing data cleaning, outlier rejection and format standardization on the water resource dynamic information, and extracting key feature data based on a distributed processing architecture; S3, analyzing the key feature data by utilizing the edge computing equipment and combining a dynamic optimization model to generate a prediction result of the water resource state, extracting feature indexes related to decision and storing the prediction result in a local storage unit; S4, generating a local management decision strategy by combining a self-adaptive algorithm with a set rule base according to the characteristic indexes acquired in real time and the stored prediction results, and transmitting the locally generated local management decision strategy to a centralized management system through a communication network; s5, verifying and adjusting the local management decision strategy by utilizing a multi-objective optimization algorithm in the centralized management system to generate a global management decision strategy meeting optimization requirements; And S6, transmitting the optimized global management decision strategy to the execution end equipment, contr