Tunnel excavation process intelligent scheduling system based on neural network

Abstract

The invention discloses a tunnel excavation process intelligent scheduling system based on a neural network, and relates to the technical field of intelligent construction management. A physical layer takes personnel, vehicles, equipment and a physical environment in tunnel construction as sensing objects; the data sensing layer acquires environmental parameters, personnel and vehicle positions and equipment states of a sensing object in real time; the data transmission layer transmits the collected parameters to the data analysis and comparison layer; the data analysis and comparison layer takes the collected parameters as input, and predicts the completion time, the completion quality and the management cost of each process of the project through a preset process content prediction and analysis model; and the function business layer monitors and adjusts the progress of each process in real time through the completion time, completion quality and management cost of each process of the project, and performs personnel and vehicle track monitoring and navigation. According to the invention, real-time data of tunnel construction can be analyzed and scheduling optimization procedures can be carried out, so that potential quality problems and construction risks can be effectively prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent construction management, and more specifically to an intelligent scheduling system for tunnel excavation processes based on neural networks. Background Art

[0002] At present, the construction management of tunnel projects relies on a Y-shaped production model that involves multiple types of work and multiple disciplines. The construction environment is complex and there are many factors that affect the construction progress, including resources, personnel, technology, environment, and funding.

[0003] Existing project management methods such as Gantt charts, network diagrams, and earned value methods can provide effective progress management at a certain stage of a single project, but lack the ability to predict and intelligently schedule the overall resource and personnel allocation of new projects. This has led to the inability of existing technologies to cope with complex tunnel construction management needs, especially the lack of intelligent analysis of real-time data and scheduling optimization, and the inability to effectively prevent potential quality problems and construction risks.

[0004] Therefore, how to analyze real-time data and perform scheduling optimization is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a neural network-based intelligent scheduling system for tunnel excavation processes to solve the problems existing in the background technology.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A neural network-based intelligent scheduling system for tunnel excavation processes, comprising: a physical layer, a data perception layer, a data transmission layer, a data analysis and comparison layer and a functional business layer, wherein the physical layer takes personnel, vehicles, equipment and physical environment in tunnel construction as perception objects; the data perception layer collects environmental parameters, personnel, vehicle positions and equipment status of the perception objects in real time; the data transmission layer transmits the environmental parameters, personnel, vehicle positions and equipment status to the data analysis and comparison layer through a transmission protocol; the data analysis and comparison layer takes the environmental parameters, personnel, vehicle positions and equipment status as inputs, and predicts the completion time, completion quality and management cost of each process of the project through a preset process content prediction and analysis model; the functional business layer monitors and adjusts the progress of each process in real time through the completion time, completion quality and management cost of each process of the project, and performs personnel and vehicle trajectory monitoring and navigation.

[0008] Preferably, it also includes a demonstration layer, which displays the progress of each process and the personnel and vehicle trajectory monitoring data uploaded by the functional business layer in real time.

[0009] Preferably, the data perception layer includes a UWB positioning base station, a sensor device and a camera.

[0010] Preferably, the transmission protocol of the data transmission layer selects any one of wireless transmission, 4 / 5G mobile communication protocol or MQTT protocol.

[0011] Preferably, the process content prediction analysis model in the data analysis comparison layer includes 9 input nodes, the first hidden layer contains 64 neurons, the activation function is ReLU, the second hidden layer contains 32 neurons, the activation function is ReLU, the output layer contains 3 output nodes, and the activation function is a linear function.

[0012] Preferably, the data analysis and comparison layer also includes a basic data processing layer for processing data transmitted by the UWB positioning base station, sensor equipment and camera.

[0013] Preferably, the functional business layer includes a process progress management module, a process cycle management module, and a project quality management module. The process progress management module monitors and adjusts the progress of each process in real time; the process cycle management module arranges and coordinates the time and sequence of each process; and the project quality management module collects various data during the tunnel excavation process in real time and generates a quality report.

[0014] Preferably, the demonstration layer includes a PC terminal, a mobile terminal and a large visual screen.

[0015] It can be seen from the above technical solution that compared with the prior art, the present invention discloses a neural network-based intelligent scheduling system for tunnel excavation processes, which reduces the waiting time in process connection through automatic scheduling and optimal resource allocation, and utilizes real-time monitoring and intelligent analysis to timely discover and correct potential quality problems, thereby realizing unattended monitoring of the entire construction cycle, reducing the burden on management personnel, and improving overall management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0017] Figure 1 A schematic diagram of the structure provided by the present invention;

[0018] Figure 2 This is a structural diagram of the model provided by the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] The embodiment of the present invention discloses a tunnel excavation process intelligent scheduling system based on neural network, such as Figure 1 As shown, it includes: a physical layer, a data perception layer, a data transmission layer, a data analysis and comparison layer, and a functional business layer. The physical layer takes personnel, vehicles, equipment, and physical environment in tunnel construction as perception objects; the data perception layer collects environmental parameters, personnel, vehicle positions, and equipment status of the perception objects in real time; the data transmission layer transmits environmental parameters, personnel, vehicle positions, and equipment status to the data analysis and comparison layer through a transmission protocol; the data analysis and comparison layer takes environmental parameters, personnel, vehicle positions, and equipment status as inputs, and predicts the completion time, completion quality, and management cost of each process of the project through a preset process content prediction and analysis model; the functional business layer monitors and adjusts the progress of each process in real time through the completion time, completion quality, and management cost of each process of the project, and conducts personnel and vehicle trajectory monitoring and navigation.

[0021] In a specific embodiment, the data perception layer collects environmental parameters, personnel and vehicle locations, and equipment status in tunnels and work areas in real time through various sensors, UWB positioning base stations, cameras, and other hardware devices.

[0022] In a specific embodiment, it also includes a demonstration layer, which displays the progress of each process and the personnel and vehicle trajectory monitoring data uploaded by the functional business layer in real time, so that the construction progress, quality status and risk warning can be known.

[0023] Among them, risk warning adopts warning notification module, including:

[0024] Risk prediction and early warning: Analyze historical data and real-time monitoring data, predict unknown risks and automatically generate early warnings to remind relevant personnel to take timely measures;

[0025] Intelligent notification system: provides flexible notification settings and automatically selects notification methods and content according to different risk levels;

[0026] Real-time incident response: When an alert occurs, an incident ticket is automatically generated and assigned to the person in charge.

[0027] and track the progress of incident handling;

[0028] Notification history record: records all warning and notification information, including triggering conditions, processing and resolution results. It is used for subsequent analysis and evaluation, optimization of warning mechanisms and response strategies, and can predict equipment failures or geological anomalies through historical data and real-time data, and provide timely warnings and adjustment plans to avoid accidents.

[0029] In a specific embodiment, the personnel and vehicle trajectory monitoring and navigation are implemented by the personnel and vehicle trajectory monitoring and navigation module, including:

[0030] Real-time positioning and monitoring: UWB positioning technology and various physical base stations and sensor devices work together to track the location of personnel and equipment during tunnel excavation in real time, providing detailed track records and map views.

[0031] Intelligent Navigation: Provides navigation functions to help personnel and equipment plan the best routes and avoid obstacles and dangerous areas during construction. Optimize navigation routes based on real-time traffic and operating conditions.

[0032] Location Reporting and Analysis: Generate detailed location information reports.

[0033] Safety management: Set up safety zones and alarm functions to monitor the safety status of personnel and equipment in real time.

[0034] Data integration and sharing: Integrate with other systems (such as scheduling system, construction management system) to support data sharing.

[0035] Customized warning rules: According to the specific needs of the project, the warning rules and processing procedures can be flexibly configured. The system supports customized risk parameters and notification settings to ensure that the warning mechanism can adapt to various complex engineering scenarios.

[0036] In a specific embodiment, the data perception layer includes a UWB positioning base station, a sensor device and a camera.

[0037] In a specific embodiment, the transmission protocol of the data transmission layer selects any one of wireless transmission (Bluetooth, LoRa, WiFi), 4 / 5G mobile communication protocol or mqtt protocol (oneNet).

[0038] In a specific embodiment, Figure 2As shown in the figure, the process content prediction and analysis model in the data analysis and comparison layer includes 9 input nodes (9 input parameters: including the average length of service of the team, team title, tunnel geological environment, development difficulty, construction urgency, process name, required equipment, process operation and other 9 parameters), the first hidden layer contains 64 neurons, the activation function is ReLU, the second hidden layer contains 32 neurons, the activation function is ReLU, the output layer contains 3 output nodes, the activation function is a linear function; the output nodes are task completion time, task completion quality and management cost. The model is trained using the Levenberg-Marquardt algorithm with 450 excavation process scheduling records as the original data set. The operation process of the prediction model is as follows: the construction management party provides project data before the excavation operation, and the model can predict the duration of each process of the project and the resource quantity summary and preset it into the system. During the construction stage, the system compares the differences in the process data uploaded in the predicted project and the actual target project (required resource quantity, completion quality, process duration), and then uploads the data and decision opinions to the functional business layer in real time.

[0039] Among them, the data analysis and comparison layer is combined with the training model, and according to the processes required for tunnel excavation, historical production data is collected and the results are analyzed. The entire chain of each process is encapsulated to establish a process preset library. With the help of the process prefabrication management function, the system parameters are adjusted according to the construction details, and the preset library automatically adjusts the library to generate a preliminary process progress model.

[0040] According to the preset rules, the production process work order is issued. During the production process, the transmission layer monitors and collects feedback from the physical layer, and the demonstration layer visualizes and compares the original process scheduling plan.

[0041] Specific steps for model building:

[0042] (1) Data preparation: A raw data set of 450 excavation process scheduling records was collected and processed, including the following fields:

[0043] Input parameters: 9 parameters including average length of service of the team, tunnel environment, process operation, etc.

[0044] Output data: task completion time, task completion quality and management cost.

[0045] The dataset is divided into training set and test set (8:2).

[0046] (2) Data preprocessing: The raw data needs to undergo a series of transformations and cleansing to ensure that it can be effectively summarized by the neural network.

[0047] Data cleaning:

[0048] To handle missing data, first check whether there are missing values ​​in the dataset, then delete the samples with missing values ​​and fill the missing values ​​with the mean value.

[0049] Remove outliers: Use box plots to visualize, observe, and remove.

[0050] Data conversion: For categorical variables in the sample, unordered categories are converted into binary variables using one-hot encoding, and for ordered categories, label encoding is used to specify integer encoding.

[0051] (3) Model definition and construction

[0052] A two-layer feedforward neural network model was built with the help of pytorch (numpy, tensorflow, scipy and keras third-party libraries were called), and the neural algorithm adopted the Levenberg-Marquardt algorithm. The model contains 9 input nodes, two hidden layers, and 3 output nodes. The input layer contains 9 input nodes, which represent 9 parameters such as the average length of service of the team, tunnel environment, and process operation. The first hidden layer contains 64 neurons, and the activation function is ReLU. The second hidden layer contains 32 neurons, and the activation function is ReLU. The output layer contains 3 output nodes, and the activation function is a linear function, which represents the task completion time, task completion quality, and management cost respectively. The model connection relationship is input node, first hidden layer, ReLU, second hidden layer, ReLU, output layer.

[0053] (4) Model training and model evaluation

[0054] Start training the model and record the losses and related indicators in the process. Use the test data set to evaluate the model performance and calculate performance indicators such as mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE). And optimize the model

[0055] (5) Model deployment and prediction

[0056] The optimized model is used for prediction to generate a preset table of tunnel excavation procedures.

[0057] In a specific embodiment, the functional business layer includes a process progress management module, a process cycle management module, and an engineering quality management module. The process progress management module monitors and adjusts the progress of each process in real time; the process cycle management module arranges and coordinates the time and sequence of each process; the engineering quality management module collects various data during the tunnel excavation process in real time and generates a quality report. The process progress management module, the process cycle management module and the engineering quality management module are in a parallel relationship.

[0058] Process progress management module module introduction:

[0059] The process progress management module focuses on real-time monitoring and adjustment of the progress of each process to ensure that the tunnel excavation project can proceed smoothly within the planned time. This module uses the output results of the process progress preset model and combines the data collected during the actual construction process to achieve dynamic adjustment and optimization.

[0060] Specific functions:

[0061] Real-time progress tracking: Through sensors and monitoring equipment, the progress data of the construction site is collected and recorded in real time, compared and analyzed with the preset model, and progress deviations are identified.

[0062] Dynamic adjustment and optimization: Based on the progress data monitored in real time, the system automatically adjusts the construction plan and displays the adjusted plan in real time on the demonstration layer.

[0063] Progress report generation: Automatically generate detailed progress reports, providing comparative analysis, progress deviations and their causes, adjustment suggestions, and expected completion time.

[0064] Resource allocation suggestions: Based on progress analysis, provide allocation suggestions for personnel, equipment and materials to ensure the best utilization of construction resources.

[0065] Process cycle management module module introduction:

[0066] The process cycle management module is dedicated to optimizing the process cycle efficiency in tunnel construction, maximizing the continuity of the construction process and reducing downtime by reasonably arranging and coordinating the time and sequence of each process.

[0067] Specific functions:

[0068] Optimize process connection: Analyze the connection between processes, identify bottlenecks in the construction process, and optimize the workflow by reordering or time adjustment.

[0069] Periodic measurement: Real-time measurement of the cycle of each process, monitoring of process time changes, and timely adjustment of process strategies to improve construction efficiency.

[0070] Periodic evaluation and feedback: Periodically evaluate the process completion status, obtain the actual completion time and quality feedback of each process, identify problems in a timely manner and take corresponding improvement measures.

[0071] Efficiency improvement strategy: Based on historical data and real-time analysis, provide strategic suggestions for improving process cycle efficiency, such as process merging, group construction and other methods.

[0072] Engineering Quality Management Module

[0073] Data collection and analysis: Real-time collection of various data during tunnel excavation, combined with the model quality parameter output of the data analysis layer, to identify potential quality problems and abnormal conditions.

[0074] Quality inspection and control: Real-time inspection of quality indicators during excavation operations, automatic generation of quality reports, detailed analysis results, and suggestions for improvement.

[0075] Abnormal warning and processing: When the system detects potential quality abnormalities or situations where personnel or vehicles deviate from predetermined standards, an early warning will be automatically triggered.

[0076] Quality Improvement Feedback: The system records all quality data and improvement measures, generating a detailed historical record.

[0077] In a specific embodiment, the presentation layer includes a PC terminal, a mobile terminal and a large visualization screen.

[0078] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0079] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A neural network-based intelligent scheduling system for tunnel excavation processes, characterized in that: include: The physical layer comprises personnel, vehicles, equipment and physical environment in tunnel construction as perception objects; the data perception layer collects environmental parameters, personnel, vehicle positions and equipment status of the perception objects in real time; the data transmission layer transmits the environmental parameters, personnel, vehicle positions and equipment status to the data analysis and comparison layer through a transmission protocol; the data analysis and comparison layer takes the environmental parameters, personnel, vehicle positions and equipment status as input, and predicts the completion time, completion quality and management cost of each process of the project through a preset process content prediction and analysis model; the functional business layer monitors and adjusts the progress of each process in real time through the completion time, completion quality and management cost of each process of the project, and conducts personnel and vehicle trajectory monitoring and navigation.

2. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 1 is characterized in that: It also includes a demonstration layer, which displays the progress of each process and the personnel and vehicle trajectory monitoring data uploaded by the functional business layer in real time.

3. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 1 is characterized in that: The data perception layer includes a UWB positioning base station, a sensor device and a camera.

4. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 1, characterized in that: The transmission protocol of the data transmission layer selects any one of wireless transmission, 4 / 5G mobile communication protocol or MQTT protocol.

5. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 3 is characterized in that: The process content prediction analysis model in the data analysis comparison layer includes 9 input nodes, the first hidden layer includes 64 neurons, the activation function is ReLU, the second hidden layer includes 32 neurons, the activation function is ReLU, the output layer includes 3 output nodes, and the activation function is a linear function.

6. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 5, characterized in that: The data analysis and comparison layer also includes a basic data processing layer for processing data transmitted by the UWB positioning base station, sensor equipment and camera.

7. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 1, characterized in that: The functional business layer includes a process progress management module, a process cycle management module, and a project quality management module. The process progress management module monitors and adjusts the progress of each process in real time; the process cycle management module arranges and coordinates the time and sequence of each process; the project quality management module collects various data during the tunnel excavation process in real time and generates a quality report.

8. The neural network-based intelligent scheduling system for tunnel excavation process according to claim 2 is characterized in that: The demonstration layer includes PC, mobile and large visual screen.

Images