A water conservancy and hydropower engineering construction management system

Abstract

The application discloses a water conservancy and hydropower engineering construction management system and relates to the technical field of water conservancy and hydropower engineering construction. In view of the problems of static management and lagging interference response of the existing system, the system comprises a multi-source data acquisition module, an interference source identification and grade evaluation module, a resource demand prediction module, a resource linkage deployment module and a warning and feedback optimization module, and forms a closed-loop management process. The system improves the accuracy of dynamic interference source identification to more than 92%, shortens the resource deployment response time by 60%, and reduces the engineering safety accident rate by 40%, and is suitable for the management of various water conservancy and hydropower engineering constructions.

Description

Technical Field

[0001] This invention relates to the field of water conservancy and hydropower engineering construction technology, and in particular to a water conservancy and hydropower engineering construction management system. Background Technology

[0002] Current water conservancy and hydropower project construction management systems mostly focus on conventional modules such as schedule planning, quality acceptance, and safety monitoring, and have the following core shortcomings: Static management of disturbance sources: The existing system can only provide passive early warning for known fixed disturbance sources (such as preset flood seasons and geological faults), and cannot capture dynamic disturbance sources that appear during construction in real time (such as temporary slope collapses, disturbances from small-scale construction projects in the surrounding area, and sudden changes in hydrological parameters).

[0003] Insufficient fusion of multi-source data: Interference source identification relies heavily on single sensor data (such as displacement sensors and water level gauges), without integrating multi-dimensional information such as drone inspection images, construction equipment operating data, and real-time meteorological data, resulting in a high misjudgment rate and delayed identification of interference sources.

[0004] Disconnect between resource allocation and interference early warning: When interference sources appear, the allocation of resources (personnel, equipment, materials) requires manual intervention and decision-making. The lack of an automatic linkage mechanism with the level and scope of interference sources leads to a lag in resource response, exacerbating project delays and cost overruns.

[0005] Existing technologies cannot solve the closed-loop management problem of "accurate identification of dynamic interference sources - quantitative assessment of interference impact - real-time linkage and allocation of resources". This deficiency is particularly prominent in water conservancy and hydropower projects in complex geological or remote areas, and there is an urgent need for a management system with dynamic early warning and intelligent linkage capabilities. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a water conservancy and hydropower project construction management system, achieving the following objectives: Real-time capture and accurate identification of dynamic interference sources during construction reduces the rate of missed or false detections of interference sources.

[0007] Quantitatively assess the impact of interference sources on project progress and quality, and generate scientific early warning information.

[0008] Based on the level of interference impact, construction resources are automatically coordinated and allocated to achieve rapid response and optimal resource allocation in dealing with interference.

[0009] The technical solution of this invention is implemented as follows: A water conservancy and hydropower project construction management system includes a multi-source data acquisition module, an interference source identification and level assessment module, a resource demand prediction module, a resource linkage and allocation module, and an early warning and feedback optimization module, which are connected sequentially. The system achieves dynamic management through the following closed-loop process: The multi-source data acquisition module collects multi-dimensional data on geology, hydrology, construction equipment conditions, environment, and images of the construction area and transmits them to the edge node for preliminary filtering. The interference source identification and level assessment module standardizes the filtered data, uses an improved DS evidence theory to fuse multi-source data features to identify dynamic interference source types, and quantifies the interference source level by constructing a comprehensive impact index. The resource demand prediction module predicts the type and quantity of resources needed to cope with the interference based on the type and level of the interference source and the current construction process, using an improved BP neural network model. The resource coordination and allocation module combines real-time resource status data, uses an improved genetic algorithm to generate and execute resource allocation plans, and tracks resource availability. The early warning and feedback optimization module issues early warning information based on the interference level and updates the model parameters of the aforementioned module based on construction feedback data.

[0010] Preferably, the slope displacement and rock stress data collected by the distributed fiber optic sensors are sampled at a frequency of 1 time / 5 minutes. The ultrasonic water level gauge and flow velocity data of the river are collected at a sampling frequency of 1 time / 10 minutes; the data of equipment load rate, running time and construction progress collected by the Internet of Things terminal of the construction equipment are collected at a sampling frequency of 1 time / 2 minutes. Wind speed and rainfall data collected by the meteorological station, and 0.1m resolution image data of the construction area collected by drones at a 2-hour / time cycle.

[0011] Preferably, the standardization process of the interference source identification and level assessment module adopts the following formula: in, The data is standardized, and x represents the original data. , These are the historical minimum and maximum values ​​for this type of data, respectively.

[0012] Preferably, the specific rules for identifying dynamic interference source types using the improved DS evidence theory in the interference source identification and level assessment module include: when the slope displacement normalization value is >0.8 and the rock mass stress normalization value is >0.75, it is identified as a "slope collapse risk interference source"; When the standardized value of rainfall is greater than 0.9 and the rise in river level is greater than 0.5 m / h, it is identified as a "flood warning interference source"; When the load rate of the construction equipment is greater than 0.95 and the continuous operating time is greater than 8 hours, and the drone image shows abnormal smoke and dust around the equipment, it is identified as an "equipment failure risk interference source".

[0013] Preferably, the formula for calculating the comprehensive impact index I constructed by the interference source identification and level assessment module is as follows: in, The safety impact coefficient is (0-1). The schedule impact factor is (0-1). The cost impact coefficient is (0-1). , , The weighting coefficients are and satisfy the following conditions: ; The interference source levels are classified according to level I as follows: Level 1 (Emergency) Level 2 (Important) Level 3 (General).

[0014] Preferably, the improved BP neural network model of the resource demand prediction module includes the type of interference source, the level of influence, and the current construction procedure as input, and the output includes the required number of personnel, the type and number of equipment, and the amount of materials used. When the input is "Level 1 slope landslide risk interference source" and the current construction procedure is "dam filling", the output is: 15 slope reinforcement personnel, 3 anchor drilling rigs, 500 high-strength anchors, and 4 emergency drainage pumps.

[0015] Preferably, the real-time resource status data of the resource linkage and allocation module is obtained through RFID tags and IoT terminals, including resource location information and availability status information; The availability status information includes whether the equipment is idle, whether the personnel are on duty, and the quantity of materials in stock.

[0016] Preferably, the objective function of the resource allocation module using an improved genetic algorithm to generate a resource allocation scheme is "shortest allocation response time + lowest allocation cost", and the constraints include that the quantity of resources meets the demand and the equipment transportation route avoids construction interference areas. When resources fail to arrive within the specified time (e.g., equipment transportation is delayed by 10 minutes), the allocation plan will be automatically adjusted, including replacing the equipment with a backup.

[0017] Preferably, the early warning information release rules of the early warning and feedback optimization module are as follows: The Level 1 warning is simultaneously issued through system audible and visual alarms, SMS messages to management personnel's mobile phones, and broadcasts at the construction site. Level 2 warnings are pushed via system messages and mobile app; Level 3 warnings are only displayed in the system background.

[0018] Preferably, the feedback optimization process of the early warning and feedback optimization module includes: construction personnel providing feedback on the interference response effect and actual resource usage through a mobile APP; Feedback data is input into the interference source identification model and the resource demand prediction model, and the confidence level of the DS evidence theory and the weight parameters of the BP neural network are updated and improved regularly.

[0019] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions: I. Overcoming existing shortcomings: Constructing a closed-loop management mechanism of "dynamic interference source identification - level assessment - resource linkage and allocation" to solve the problems of static management and delayed response in the existing system.

[0020] II. Improved Identification Accuracy: Through multi-source data fusion and improved DS evidence theory, the accuracy of dynamic interference source identification has been improved to over 92%, which is 35% higher than that of single data identification methods.

[0021] III. Resource Allocation Optimization: Resource allocation response time is shortened to within 30 minutes, improving efficiency by 60% compared to manual allocation, while reducing resource waste rate by more than 25%.

[0022] IV. Reduced Engineering Risks: Through dynamic early warning and rapid response, the incidence of safety accidents caused by interference sources during engineering construction is reduced by 40%, and the project delay rate is reduced by 30%.

[0023] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a diagram of the overall system architecture of the present invention; Figure 2 This is a system closed-loop flowchart of the present invention; Figure 3 This is an interactive diagram of interference source identification and resource allocation in this invention. Detailed Implementation

[0026] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0027] It is important to note that terms such as "first," "second," "symmetric," and "array" are used only to distinguish between descriptive and positional descriptions and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features specified with terms such as "first" or "symmetric" may explicitly or implicitly include one or more of that feature; similarly, when the quantity of certain features is not limited by words such as "two" or "three," it should be noted that such features also explicitly or implicitly include one or more features. In this invention, unless otherwise explicitly specified and limited, terms such as "installation," "connection," and "fixation" should be interpreted broadly; for example, they can refer to a fixed connection, a detachable connection, or an integral molding; they can refer to a mechanical connection, a direct connection, a welding connection, or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the accompanying drawings and specific circumstances.

[0028] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] like Figure 1-3 The present invention provides a water conservancy and hydropower project construction management system, comprising a multi-source data acquisition module, an interference source identification and level assessment module, a resource demand prediction module, a resource linkage and allocation module, and an early warning and feedback optimization module, which are connected sequentially. The system achieves dynamic management through the following closed-loop process: Each module achieves high-speed data flow and preliminary processing through the 5G+ edge computing architecture. The edge nodes filter noise and remove outliers from the collected data before uploading it to the system cloud platform to ensure data quality and transmission efficiency. The multi-source data acquisition module collects multi-dimensional data on geology, hydrology, construction equipment conditions, environment, and images of the construction area and transmits them to edge nodes for preliminary filtering. Among them, the distributed fiber optic sensor adopts a density of one monitoring point every 50 meters to achieve millimeter-level precision measurement of slope displacement and rock mass stress; The drone is equipped with a multispectral imaging lens and performs inspection missions every 2 hours according to the route plan of "covering the entire construction area + increasing the density in key areas"; The acquisition of multi-dimensional and high-precision raw data provides a comprehensive and accurate basis for subsequent interference source identification. The pre-processing of edge computing effectively reduces the computing load of the cloud platform, improving the overall system response speed by 40%.

[0030] The standardized processing of the interference source identification and level assessment module adopts the following formula: in, The data is standardized, and x represents the original data. , These are the historical minimum and maximum values ​​for this type of data; for different types of data such as slope displacement and rainfall, the system establishes independent historical extreme value databases to ensure the relevance of the standardization process. At the same time, the standardized data is validated a second time. If the data exceeds the range of [0,1], the abnormal data marking mechanism is triggered. The interference source identification and level assessment module uses the improved DS evidence theory to identify the specific rules for dynamic interference source types, including: when the slope displacement normalization value is >0.8 and the rock mass stress normalization value is >0.75, it is identified as a "slope collapse risk interference source"; When the standardized value of rainfall is greater than 0.9 and the rise in river level is greater than 0.5 m / h, it is identified as a "flood warning interference source"; When the load rate of construction equipment is greater than 0.95 and the continuous operating time is greater than 8 hours, and the drone image shows abnormal smoke and dust around the equipment, it is identified as a "source of equipment failure risk interference". The improved DS evidence theory introduces a dynamic weight adjustment mechanism, which adjusts the evidence synthesis weights according to the real-time credibility of different data sources (such as sensor online rate and data volatility), thus solving the decision bias problem of traditional DS theory when evidence conflicts occur. Standardization eliminates the dimensional differences of multi-source data, enabling different types of data to be fused and analyzed in the same dimension; the improved application of DS evidence theory increases the accuracy of dynamic interference source identification to over 92%, and reduces the false judgment rate by 60% compared to traditional single-data identification methods.

[0031] The formula for calculating the comprehensive impact index I constructed by the interference source identification and level assessment module is as follows: in, The safety impact factor is (0-1). The schedule impact factor is (0-1). The cost impact coefficient is (0-1). , , The weighting coefficients are and satisfy the following conditions: ; Weighting coefficient , , It can be based on the project stage (such as peak construction period) Increased to 0.5, during the acceptance phase. Increased to 0.6), types of disturbance sources (such as slope collapse risk) Set to 0.5, flood warning (Set to 0.4) for dynamic configuration, automatically optimized through the system's built-in weight adjustment algorithm; Interference source levels are classified as follows (I): Level 1 (Emergency) Level 2 (Important) Level 3 (General); Dynamic weighting adjustments make the comprehensive impact index more closely aligned with actual engineering scenarios, improving the scientific rigor of quantitative assessment results by 30%, providing precise priority criteria for resource allocation, and avoiding blind resource investment.

[0032] The improved BP neural network model for the resource demand forecasting module takes into account the type of interference source, the level of impact, and the current construction procedure as inputs, and outputs the required number of personnel, equipment type and quantity, and material usage. When the input is "Level 1 slope collapse risk interference source" and the current construction procedure is "dam filling", the output is: 15 slope reinforcement personnel, 3 anchor drilling rigs, 500 high-strength anchors, and 4 emergency drainage pumps. The improved BP neural network introduces an attention mechanism layer, which assigns higher weights to key input features such as interference level and construction procedure. At the same time, batch normalization technology is used to accelerate model convergence, reducing the resource prediction time to less than 10 seconds. The real-time resource status data of the resource linkage and allocation module is obtained through RFID tags and IoT terminals, including resource location information and availability status information; Available status information includes whether the equipment is idle, whether the personnel are on duty, and the quantity of materials in stock; The RFID tags use the ultra-high frequency band, with a reading distance of up to 10 meters, and support the simultaneous identification of multiple tags. The IoT terminal is based on the LoRaWAN protocol to achieve low-power wide-area communication, ensuring real-time transmission of resource status in remote construction areas. The resource coordination and allocation module uses an improved genetic algorithm to generate resource allocation schemes. The objective function is "shortest allocation response time + lowest allocation cost". The constraints include meeting the demand for resources and ensuring that equipment transportation routes avoid construction interference areas. When resources fail to arrive on time (e.g., equipment transportation is delayed by 10 minutes), the allocation plan will be automatically adjusted, including replacing the equipment with a backup. The improved genetic algorithm introduces an elite retention strategy in the selection operator and adopts adaptive crossover probability in the crossover operator, which improves the search efficiency of the optimal allocation scheme by 50%. Improved BP neural networks increased the accuracy of resource demand prediction to 95%, reducing resource waste; improved combination of genetic algorithms and IoT technology shortened the resource allocation response time to within 30 minutes, improving efficiency by 60% compared to manual allocation, and effectively reducing the impact of interference sources on project progress.

[0033] The rules for issuing early warning information in the early warning and feedback optimization module are as follows: The Level 1 warning is simultaneously issued through system audible and visual alarms, SMS messages to management personnel's mobile phones, and broadcasts at the construction site. The system's audible and visual alarm uses an industrial-grade integrated audible and visual device, with an alarm sound pressure level ≥120dB and a visual distance of ≥500 meters for the optical signal; Management personnel's mobile phone text messages are sent via the highest priority channel to ensure delivery within 1 minute; The construction site broadcast coverage radius is ≥1 kilometer, and it supports independent broadcasting in multiple zones; Level 2 warnings are pushed via system messages and mobile app; Level 3 warnings are only displayed in the system background. The feedback optimization process of the early warning and feedback optimization module includes: construction personnel providing feedback on the effectiveness of interference response and the actual use of resources through a mobile app; The feedback data is input into the interference source identification model and the resource demand prediction model, and the confidence of the DS evidence theory and the weight parameters of the BP neural network are updated and improved regularly. The feedback data is updated in real time using an online learning algorithm. Every 100 valid feedbacks are received triggers an iterative optimization of the model, ensuring that the system's accuracy continues to improve over the long term. The multi-channel, high-priority early warning dissemination ensured a 100% reach rate for Level 1 early warnings, guaranteeing timely response to interference. The closed-loop feedback optimization mechanism enables the system to have self-learning capabilities, and the accuracy of the recognition and prediction models continues to improve as the project progresses, achieving the effect of "becoming smarter the more it is used".

[0034] In this embodiment, the present invention operates as follows: First, multi-source data acquisition is conducted: slope displacement and rock stress data are collected every 5 minutes using distributed fiber optic sensors; river water level and flow velocity data are collected every 10 minutes using ultrasonic level gauges and flow velocity meters; equipment load rate, runtime, and construction progress data are collected every 2 minutes using IoT terminals on construction equipment; wind speed and rainfall data are collected in real time by a weather station; and image data of the construction area is collected every 2 hours by drones at a resolution of 0.1m. All collected data is transmitted to edge nodes for preliminary filtering to remove noise and abnormal data before being uploaded to the system cloud platform. Then, interference source identification and level assessment were performed: the filtered data were processed using a formula... After standardization, the improved DS evidence theory is used to fuse multi-source data features to identify the type of dynamic disturbance source. For example, when the standardized value of slope displacement > 0.8 and the standardized value of rock mass stress > 0.75, it is identified as a "slope collapse risk disturbance source". Then, the formula is used... Calculate the comprehensive impact index I, and according to... , , Interference sources are classified into three levels: Level 1 (urgent), Level 2 (important), and Level 3 (general). Then, resource demand prediction is performed: the type and level of the interference source and the current construction procedure are input into the improved BP neural network model. For example, when the input is "Level 1 slope collapse risk interference source" and the current construction procedure is "dam filling", the output is the resource demand of 15 slope reinforcement personnel, 3 anchor drilling rigs, 500 high-strength anchors, and 4 emergency drainage pumps. Next, resource coordination and allocation are carried out: real-time data such as resource location and availability (equipment idleness, personnel on duty, and material inventory) are obtained through RFID tags and IoT terminals. An improved genetic algorithm is used to generate and execute an allocation plan with the goal of "shortest allocation response time and lowest allocation cost" in combination with constraints such as resource quantity meeting demand and equipment transportation routes avoiding construction interference areas. At the same time, the arrival of resources is tracked. If equipment transportation is delayed by 10 minutes, the plan is automatically adjusted (such as replacing with backup equipment). Finally, early warning and feedback optimization: Early warning information is issued based on the interference level. Level 1 warnings are simultaneously issued via system audible and visual alarms, SMS messages to management personnel, and on-site broadcasts; Level 2 warnings are pushed via system messages and a mobile app; Level 3 warnings are only displayed in the system background. Construction personnel provide feedback on the effectiveness of interference response and actual resource usage via the mobile app. The system inputs this feedback data into the interference source identification model and resource demand prediction model, regularly updating and improving the confidence level of the DS evidence theory and the weight parameters of the BP neural network to achieve continuous model optimization.

[0035] The following are several other specific embodiments of the application of this invention: Example 1: An Example of a Large-Scale Water Conservancy Project First, multi-source data acquisition was conducted: 20 distributed fiber optic sensors were deployed at the dam foundation and high slopes on both banks, collecting slope displacement and rock stress data every 5 minutes; 3 sets of ultrasonic level gauges and flow velocity meters were deployed in the upstream and downstream channels of the project, collecting river level and flow velocity data every 10 minutes; IoT terminals were installed on more than 100 construction equipment, including tunnel boring machines and large cranes, collecting equipment load rate, runtime, and construction progress data every 2 minutes; 3 weather stations were deployed to collect wind speed and rainfall in real time, and drones inspected the construction area every hour, improving image resolution to 0.05m. All data was transmitted to edge nodes via a dual-link "5G + satellite" system to ensure uninterrupted data transmission in remote areas. Then, interference source identification and level assessment were performed: the filtered data were processed using a formula... Standardization was performed, and then multi-source data features were integrated using an improved DS evidence theory. When the standardized value of the dam foundation rock mass stress > 0.85 and the standardized value of the slope displacement > 0.9, it was identified as a "source of dam foundation instability risk interference"; when the daily rise in upstream water level > 2m and the standardized value of rainfall > 0.95, it was identified as a "source of interference for severe flood warnings". A comprehensive impact index was constructed. If a "source of dam foundation instability risk" is identified and it occurs during a critical stage of the dam's concrete pouring process, substitute it with... , , ,have to It was determined to be a Level 1 warning; Then, resource demand forecasting was performed: the "first-level dam foundation instability risk interference source" and "concrete pouring process" were input into the improved BP neural network model, and the output resource demand was: 20 geological reinforcement personnel, 5 hydraulic drilling rigs, 800 high-strength anchor cables, and 20 tons of emergency leak-sealing materials. Next, resource coordination and allocation were implemented: Real-time monitoring of resource status via RFID tags and IoT terminals revealed 3 idle hydraulic drilling rigs (located in a temporary work area 5km away), 15 on-duty geological personnel, and 600 anchor cables in stock (200 of which needed to be transported from a material yard 20km away). An improved genetic algorithm was used to plan transportation routes with the goal of "shortest allocation response time + lowest allocation cost," prioritizing the dispatch of the 3 drilling rigs in the work area (transport time 20 minutes). Simultaneously, 5 backup geological personnel were notified to report to their posts, and the anchor cables were expedited by dedicated vehicles from the material yard (arriving within 30 minutes). Finally, there is the optimization of early warning and feedback: After a Level 1 early warning is triggered, the system simultaneously broadcasts an audible and visual alarm, sends SMS messages to management personnel, and broadcasts the information in three broadcast zones at the construction site. After construction is completed, feedback indicates that the actual consumption of anchor cables was 180 tons and the drilling rig utilization rate was 100%. The system inputs the feedback data into the model to update and improve the confidence weight of "dam foundation instability" in the DS evidence theory, thereby optimizing the accuracy of subsequent identification.

[0036] Example 2: Example of a hydropower station project in a remote area First, multi-source data acquisition was conducted: eight distributed fiber optic sensors were deployed on the slopes surrounding the power plant to collect slope displacement and rock stress data every 5 minutes; a hydrological monitoring system was deployed in the river channel to collect water level and flow velocity data every 15 minutes (the sampling frequency was appropriately reduced due to communication bandwidth limitations in remote areas); IoT terminals were installed on 15 construction equipment such as excavators and loaders to collect equipment operating data every 5 minutes; a weather station was deployed, and drones inspected the construction area every 3 hours. The collected data was initially stored on edge nodes and uploaded to the cloud in batches every hour. Subsequently, interference source identification and level assessment were conducted: After data standardization, an improved DS evidence theory was used to identify interference sources. When the standardized slope displacement value > 0.8 and the standardized rainfall value > 0.85, it was identified as a "slope landslide risk interference source"; when the construction equipment load rate > 0.98 and the continuous operating time > 10 hours, and the UAV imagery showed overheating equipment, it was identified as an "equipment failure risk interference source". A comprehensive impact index was constructed. If a "source of equipment failure risk" is identified and it is located during a critical stage of unit installation, substitute it into... , , ,have to It was determined to be a Level 1 warning; Then, resource demand forecasting is performed: "Level 1 equipment failure risk interference source" and "unit installation process" are input into the improved BP neural network model, and the output resource demand is: 8 equipment maintenance personnel, 1 standby generator, and 2 sets of maintenance tools. Next, resource coordination and allocation were carried out: due to the scarcity of resources in remote areas, the resource coordination and allocation module prioritized the use of on-site backup resources (the backup generator is located in the factory warehouse and will arrive in 5 minutes). Maintenance personnel coordinated with nearby work areas through on-site duty (arriving in 10 minutes). At the same time, the remote expert support system was activated to provide video guidance for maintenance. Finally, there's the optimization of early warning and feedback: Level 1 early warnings are issued via system audible and visual alarms, satellite phone text messages for management personnel, and on-site broadcasts. After maintenance is completed, feedback indicates that maintenance time is 20% shorter than predicted. The system then inputs the feedback data into the model to update and improve the weight parameters of the BP neural network, thereby enhancing the accuracy of subsequent equipment failure resource predictions.

[0037] Example 3: Implementation of a Flood Control Project in a Certain City First, multi-source data collection was conducted: 15 distributed fiber optic sensors were deployed along the flood control dike to collect dike displacement data every 5 minutes; 5 sets of hydrological monitoring equipment were deployed at the confluence of urban rivers and external rivers to collect water level and flow velocity data every 5 minutes (urban hydrology changes rapidly, so the sampling frequency was increased); IoT terminals were installed on 25 small excavators, pile drivers and other construction equipment to collect equipment operating data every 2 minutes; 3 meteorological stations were deployed, and drones patrolled every 1.5 hours, focusing on monitoring image data of municipal construction and building projects around the flood control dike; Subsequently, interference sources were identified and their levels assessed: After data standardization, an improved DS evidence theory was used to identify interference sources. When the standardized value of the dike displacement > 0.75 and the standardized value of rainfall > 0.9, it was identified as a "flood control dike breach risk interference source"; when the distance of surrounding construction equipment from the flood control dike is < 5m and the equipment load rate is > 0.9, and UAV imagery shows earthwork disturbance at the dike toe, it was identified as a "surrounding construction disturbance interference source". A comprehensive impact index was constructed. If a "source of disturbance from surrounding construction" is identified and it occurs during a critical phase of flood control dike reinforcement, substitute it into... , , ,have to It was determined to be a Level 1 warning; Then, resource demand forecasting was performed: the "primary surrounding construction disturbance sources" and the "flood control dike reinforcement process" were input into the improved BP neural network model, and the output resource demand was: 12 people for dike reinforcement, 2 hydraulic pile drivers, and 500㎡ of impermeable geomembrane. Next, resource coordination and allocation were carried out: resource status was monitored through RFID tags and IoT terminals, revealing one idle hydraulic pile driver (located 3km away from the work area, with a transportation time of 15 minutes) and nine on-site reinforcement personnel (three of whom needed to be transferred from nearby teams). The genetic algorithm was improved to plan routes that avoided congested urban traffic sections, ensuring the pile driver arrived within 20 minutes and personnel assembled within 10 minutes. Finally, the early warning and feedback optimization: Level 1 early warnings are simultaneously released via system audible and visual alarms, a mobile app for management personnel, and the large screen at the flood control command center. After reinforcement was completed, feedback indicated that 480 square meters of geomembrane were actually used. The system inputs this feedback data into the model to update and improve the identification rules for "surrounding construction disturbance" in the DS evidence theory, thereby enhancing the accuracy of interference source identification in urban scenarios.

[0038] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A water conservancy and hydropower project construction management system, characterized in that, The system comprises a multi-source data acquisition module, an interference source identification and level assessment module, a resource demand prediction module, a resource linkage and allocation module, and an early warning and feedback optimization module, all connected sequentially. The system achieves dynamic management through the following closed-loop process: The multi-source data acquisition module collects multi-dimensional data on geology, hydrology, construction equipment conditions, environment, and images of the construction area and transmits them to the edge node for preliminary filtering. The interference source identification and level assessment module standardizes the filtered data, uses an improved DS evidence theory to fuse multi-source data features to identify dynamic interference source types, and quantifies the interference source level by constructing a comprehensive impact index. The resource demand prediction module predicts the type and quantity of resources needed to cope with the interference based on the type and level of the interference source and the current construction process, using an improved BP neural network model. The resource coordination and allocation module combines real-time resource status data, uses an improved genetic algorithm to generate and execute resource allocation plans, and tracks resource availability. The early warning and feedback optimization module issues early warning information based on the interference level and updates the model parameters of the aforementioned module based on construction feedback data.

2. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The slope displacement and rock stress data were collected by distributed fiber optic sensors at a sampling frequency of 1 time / 5 minutes. The ultrasonic water level gauge and flow velocity data of the river are collected at a sampling frequency of 1 time / 10 minutes; the data of equipment load rate, running time and construction progress collected by the Internet of Things terminal of the construction equipment are collected at a sampling frequency of 1 time / 2 minutes. Wind speed and rainfall data collected by the meteorological station, and 0.1m resolution image data of the construction area collected by drones at a 2-hour / time cycle.

3. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The standardized processing of the interference source identification and level assessment module adopts the following formula: in, The data is standardized, and x represents the original data. , These are the historical minimum and maximum values ​​for this type of data, respectively.

4. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The interference source identification and level assessment module uses the improved DS evidence theory to identify the specific rules for dynamic interference source types, including: when the slope displacement normalization value is >0.8 and the rock mass stress normalization value is >0.75, it is identified as a "slope collapse risk interference source"; When the standardized value of rainfall is greater than 0.9 and the rise in river level is greater than 0.5 m / h, it is identified as a "flood warning interference source"; When the load rate of the construction equipment is greater than 0.95 and the continuous operating time is greater than 8 hours, and the drone image shows abnormal smoke and dust around the equipment, it is identified as an "equipment failure risk interference source".

5. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The formula for calculating the comprehensive impact index I constructed by the interference source identification and level assessment module is as follows: in, The safety impact coefficient is (0-1). The schedule impact factor is (0-1). The cost impact coefficient is (0-1). , , The weighting coefficients are and satisfy the following conditions: ; The interference source levels are classified according to level I as follows: Level 1 (Emergency) Level 2 (Important) Level 3 (General).

6. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The improved BP neural network model of the resource demand prediction module takes into input the type of interference source, the level of influence, and the current construction procedure, and outputs the required number of personnel, equipment type and number of units, and material usage. When the input is "Level 1 slope landslide risk interference source" and the current construction procedure is "dam filling", the output is: 15 slope reinforcement personnel, 3 anchor drilling rigs, 500 high-strength anchors, and 4 emergency drainage pumps.

7. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The real-time resource status data of the resource linkage and allocation module is obtained through RFID tags and IoT terminals, including resource location information and availability status information; The availability status information includes whether the equipment is idle, whether the personnel are on duty, and the quantity of materials in stock.

8. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The resource coordination and allocation module uses an improved genetic algorithm to generate resource allocation schemes with the objective function of "shortest allocation response time + lowest allocation cost". The constraints include that the quantity of resources meets the demand and the equipment transportation route avoids construction interference areas. When resources fail to arrive within the specified time (e.g., equipment transportation is delayed by 10 minutes), the allocation plan will be automatically adjusted, including replacing the equipment with a backup.

9. The water conservancy and hydropower project construction management system according to claim 1, characterized in that: The early warning information release rules of the early warning and feedback optimization module are as follows: The Level 1 warning is simultaneously issued through system audible and visual alarms, SMS messages to management personnel's mobile phones, and broadcasts at the construction site. Level 2 warnings are pushed via system messages and mobile app; Level 3 warnings are only displayed in the system background.

10. A water conservancy and hydropower project construction management system according to claim 1, characterized in that: The feedback optimization process of the early warning and feedback optimization module includes: construction personnel providing feedback on the interference response effect and actual resource usage through a mobile APP; Feedback data is input into the interference source identification model and the resource demand prediction model, and the confidence level of the DS evidence theory and the weight parameters of the BP neural network are updated and improved regularly.

Images