Truck anti-lifting control method and device based on lane condition factor analysis

Abstract

The application provides a container truck anti-lifting control method and device based on lane condition factor analysis: a laser scanner is used to collect container ranging data in the lifting-lug box stage, a camera synchronously captures multi-lane vehicle information, and multi-source input data is generated; a double-factor load model is constructed according to the multi-source input data, a first factor is a real ranging component of a working lane, and a second factor is an adjacent overtaking lane interference component; a model covariance matrix is decomposed into a load matrix and an independent noise matrix, the load matrix is orthogonally rotated, and the load components of the working lane and the adjacent overtaking lane interference are separated; the separated ranging value of the working lane is calculated based on the rotated matrix, compared with the initial ranging value, and when the difference exceeds a threshold value, a brake instruction is generated; the instruction is transmitted to a rail crane control system, and the lifting of the lifting-lug is interrupted. Through multi-sensor data fusion and covariance calculation, the method decouples the interference, accurately estimates the distance, brakes in time, improves the accuracy of the anti-lifting judgment, and ensures the safety and efficiency of the operation.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of container truck anti-lifting, and particularly relates to a container truck anti-lifting control method and device based on lane condition factor analysis. BACKGROUND

[0002] In the field of automatic operation control technology of port container gantry cranes, with the continuous growth of port throughput and the continuous improvement of automation level, the bridge crane (referred to as "track crane") plays an increasingly important role in port production. The widespread application of fully automated container loading and unloading processes has greatly improved the production efficiency of the port. However, in the process of automated port transportation, ensuring the safe separation between the container and the trailer board of the container truck has become a crucial task. The traditional container truck anti-lifting system often misjudges in dealing with complex and variable operating environments, especially in multi-lane operations and when vehicles in adjacent lanes overtake, that is, the container and the trailer board in the unloading operation are incorrectly judged as not separated due to the passing of other lane container trucks, and then the lifting of the spreader is stopped, which seriously affects the efficiency of container loading and unloading operations.

[0003] Specifically, the traditional system usually relies on the data of a single sensor or a limited number of sensors for judgment, which may include laser scanners, cameras, etc. However, in a multi-lane operating environment, due to the influence of various factors such as vehicle activity in adjacent lanes, changes in lighting conditions, obstructions, etc., the data of a single sensor often has uncertainty, redundancy, or even contradictions, making it difficult to accurately reflect the actual contact state between the container and the trailer board of the container truck. For example, the laser scanner may be disturbed by vehicles in adjacent lanes during scanning, resulting in deviation in distance measurement data; and the camera may not be able to clearly capture the separation of the container and the trailer board due to insufficient light, obstruction, etc. SUMMARY

[0004] The purpose of the present application is to overcome the defects in the prior art and provide a container truck anti-lifting control method and device based on lane condition factor analysis.

[0005] A container truck anti-lifting control method based on lane condition factor analysis is provided, comprising:

[0006] Collecting container distance measurement data at consecutive time points during the lifting of the spreader by a laser scanner, and synchronously capturing multi-lane vehicle information by a camera to generate multi-source input data;

[0007] Building a double-factor load model according to the multi-source input data, with the first factor being the real distance measurement component of the operating lane and the second factor being the interference component of the adjacent overtaking lane;

[0008] Decomposing the covariance matrix of the double-factor load model into a load matrix and an independent noise matrix;

[0009] performing orthogonal rotation on the load matrix to separate the load components of the working lane from the adjacent overtaking lane;

[0010] calculating a working lane separation ranging value based on the rotated load matrix and the noise matrix;

[0011] comparing the working lane separation ranging value with an initial ranging value, and generating a braking instruction when the difference exceeds a preset threshold;

[0012] transmitting the braking instruction to a rail-mounted crane control system to execute a spreader hoisting interruption action.

[0013] Optionally, the laser scanner is installed in the middle region of the rail-mounted crane trolley, and the scanning direction is parallel to the truck lane.

[0014] The camera is installed in the middle of the rail-mounted crane jib, covering the working lane below the jib.

[0015] The camera is installed in the middle of the rail-mounted crane jib, covering the working lane below the jib.

[0016] Optionally, transmitting the braking instruction to a rail-mounted crane control system to execute a spreader hoisting interruption action, including:

[0017] The braking instruction is sent to the rail-mounted crane control system by wired or wireless transmission, and is synchronously transmitted to a remote monitoring terminal.

[0018] Optionally, after calculating a working lane separation ranging value based on the rotated load matrix and the noise matrix, including:

[0019] When it is detected that the overtaking lane interference component load value exceeds a second preset threshold, the number of time points of the multi-source input data is expanded and the working lane separation ranging value is recalculated.

[0020] Optionally, before generating multi-source input data by collecting container ranging data at continuous time points during the spreader container loading stage by a laser scanner and synchronously capturing multi-lane vehicle information by a camera, including:

[0021] When the rail-mounted crane controller is started, the double-factor load model is deployed to an intelligent analysis server, and real-time communication is established with the laser scanner and the camera.

[0022] The application also provides a truck anti-slinging control device based on lane condition factor analysis, including:

[0023] The acquisition module acquires container ranging data at continuous time points during the stage of the spreader setting the container through a laser scanner, synchronously captures multi-lane vehicle information through a camera, and generates multi-source input data.

[0024] The model module constructs a double-factor load model according to the multi-source input data, a first factor being a real ranging component of a working lane and a second factor being an adjacent overtaking lane interference component.

[0025] The decomposition module decomposes a covariance matrix of the double-factor load model into a load matrix and an independent noise matrix.

[0026] The rotation module implements orthogonal rotation on the load matrix to separate the load components of the working lane and the adjacent overtaking lane interference.

[0027] The calculation module calculates a working lane separated ranging value based on the rotated load matrix and the noise matrix.

[0028] The comparison module compares the working lane separated ranging value with an initial ranging value, and generates a braking instruction when a difference exceeds a preset threshold.

[0029] The execution module transmits the braking instruction to a rail-mounted crane control system to execute a spreader hoisting interruption action.

[0030] Optionally, the acquisition module acquires container ranging data at continuous time points during the stage of the spreader setting the container through a laser scanner, synchronously captures multi-lane vehicle information through a camera, and generates multi-source input data, including:

[0031] The laser scanner is installed in a middle region of a rail-mounted crane trolley, and a scanning direction is parallel to a truck lane.

[0032] The camera is installed in a middle region of a rail-mounted crane cantilever, and covers a working lane below the cantilever.

[0033] Optionally, the execution module transmits the braking instruction to a rail-mounted crane control system to execute a spreader hoisting interruption action, including:

[0034] The braking instruction is sent to the rail-mounted crane control system through a wired or wireless transmission mode, and is synchronously transmitted to a remote monitoring terminal.

[0035] Optionally, after the calculation module calculates a working lane separated ranging value based on the rotated load matrix and the noise matrix, including:

[0036] When it is detected that an overtaking lane interference component load value exceeds a second preset threshold, the number of time points of the multi-source input data is expanded, and a working lane separated ranging value is recalculated.

[0037] Optionally, the collection module collects container ranging data at continuous time points during the stage of the spreader setting the container through a laser scanner, and synchronously captures multi-lane vehicle information through a camera, before generating multi-source input data, comprising:

[0038] When the rail-mounted crane controller is started, the double-factor load model is deployed to the intelligent analysis server and real-time communication is established with the laser scanner and the camera.

[0039] The application has the following beneficial effects:

[0040] The application provides a container truck anti-lifting control method based on lane condition factor analysis, comprising: collecting container ranging data at continuous time points during the stage of the spreader setting the container through a laser scanner, and synchronously capturing multi-lane vehicle information through a camera, to generate multi-source input data; constructing a double-factor load model according to the multi-source input data, the first factor being a real ranging component of the working lane, and the second factor being an adjacent overtaking lane interference component; decomposing the covariance matrix of the double-factor load model into a load matrix and an independent noise matrix; performing orthogonal rotation on the load matrix to separate the load components of the working lane and the adjacent overtaking lane interference; calculating a working lane separated ranging value based on the rotated load matrix and the noise matrix; comparing the working lane separated ranging value with an initial ranging value, and generating a brake command when the difference exceeds a preset threshold; and transmitting the brake command to a rail-mounted crane control system to execute a spreader lifting interruption action. Through multi-sensor data fusion and covariance calculation, multi-lane interference is decoupled, the distance between the container and the container truck deck is accurately estimated, a brake signal is generated in time, the accuracy of the anti-lifting judgment is improved, and the work safety and efficiency are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 is a container truck anti-lifting control flowchart based on lane condition factor analysis. DETAILED DESCRIPTION

[0042] Exemplary embodiments of the present disclosure will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it is understood that various forms implement the present disclosure and should not be limited by the embodiments set forth herein. Rather, the embodiments are provided to enable a more thorough understanding of the present disclosure and to convey the full scope of the present disclosure to those skilled in the art.

[0043] The application proposes a container truck anti-lifting system based on multi-sensor data fusion technology. The system corrects and allocates data weights to all participating sensors through multi-sensor data fusion technology, i.e., multi-sensor information fusion technology, eliminates possible redundancies and contradictions between multi-sensor information, mutually supplements, and improves system accuracy together.

[0044] The technical solution of the present application is that the laser scanner is installed in the possible influence area of the middle part of the rail-mounted crane trolley, which is used to scan the horizontal distance in the direction of the truck lane; the camera is arranged in the middle part of the cantilever of the rail-mounted crane, which is used to collect the information of the vehicles in multiple lanes under the cantilever; the intelligent analysis server as the core computer of the system has three functions:

[0045] (1) The image processing technology is adopted to perform image preprocessing on the multiple-lane information collected by the camera.

[0046] (2) The data fusion technology is adopted to perform credibility determination and weight distribution on the information of multiple sensors.

[0047] (3) The control logic judgment function of the whole system is borne.

[0048] The network transmission device is responsible for collecting the control instructions of the intelligent analysis server and transmitting the instructions to the control system; the control system can be located in the operation room or the remote control center, and the control system is the final execution mechanism of the user operation device and the system control. Further, the laser scanner starts the data collection action after the completion of the container lifting by the spreader during the unloading operation, and establishes the initial data before lifting.

[0049] Please refer to Figure 1 , the present application provides a truck anti-lifting control method based on lane condition factor analysis, which comprises:

[0050] S101, collecting the container ranging data at continuous time points during the container lifting by the spreader through the laser scanner, synchronously capturing the multiple-lane vehicle information through the camera, and generating multiple-source input data;

[0051] When the container is lifted from the truck by locking, the laser scanner detects the gap between the container and the truck pallet in real time, and collects real-time data. At the same time, the camera installed in the cantilever collects the real-time state of each lane.

[0052] The real-time data collected by the laser scanner and the real-time state of each lane collected by the camera are taken as the input conditions of the fusion algorithm for data fusion, and the confidence of the real-time data of the laser scanner is updated, so as to obtain the corrected data after the fusion of multiple sensors. Through the comparison between the corrected data and the initial data, the judgment operation of whether the container is separated from the truck pallet is performed.

[0053] S102, constructing a double-factor load model according to the multiple-source input data, the first factor being a real ranging component of the working lane, and the second factor being an adjacent overtaking lane interference component;

[0054] Λ: factor load matrix, describing the sensitivity of ranging data to the working lane and the overtaking lane.

[0055] : latent factors, satisfying , .

[0056] : independent measurement noise, , (diagonal matrix).

[0057] Decompose the covariance matrix:

[0058] Covariance matrix of observed variables:

[0059]

[0060]

[0061] and independent)

[0062]

[0063] S103、Decompose the covariance matrix of the double-factor loading model into a loading matrix and an independent noise matrix;

[0064] Assume that the observed data follows a multivariate normal distribution , the log-likelihood function is:

[0065]

[0066] where, is the sample covariance matrix.

[0067] Parameter estimation steps:

[0068] Initialization: Initialize the factor loading matrix Λ using principal component analysis (PCA):

[0069] and ,

[0070] where , are the first two largest eigenvalues, , are the corresponding eigenvectors.

[0071] Iterative optimization:

[0072] Fixing Ψ, update Λ: .

[0073] Fixing Λ, update Ψ: .

[0074] Convergence condition: stop when the parameter change is less than a threshold (such as ).

[0075] S104, orthogonal rotation is performed on the load matrix to separate the load components of the working lane and the adjacent overtaking lane interference;

[0076] To enhance the interpretability of the factor load, the orthogonal rotation is performed on Λ:

[0077]

[0078] Wherein, the rotation matrix Maximizing the variance of the square of the factor load ensures that the working lane factor ( ) mainly reflects the ranging signal of the target lane, and the adjacent overtaking lane factor ( ) mainly reflects the interference signal.

[0079] Extract the real data of the working lane:

[0080] Calculate the factor score by regression method Estimate the real ranging value of the working lane:

[0081]

[0082] S105, based on the rotated load matrix and the noise matrix, calculate the working lane separation ranging value;

[0083] Suppose that the ranging data at a certain time is affected by the following factors:

[0084]

[0085] Covariance matrix generation:

[0086]

[0087]

[0088] Factor score extraction:

[0089] Suppose that the observation data is Calculate the working lane factor score:

[0090]

[0091] Wherein, is the load column corresponding to the working lane.

[0092] The output results are tested and normalized. The above steps are repeated until the optimal solution is obtained.

[0093] Further, the control instruction issued by the intelligent analysis server is transmitted to the control system through the network transmission device to perform the corresponding control action.

[0094] Furthermore, the control instructions are electric control instructions implemented by combining the button signals of the electric control box and the PLC program logic.

[0095] Furthermore, the control instructions sent by the intelligent analysis server are directly sent to the PLC program logic to change the state of the electric control button to achieve the control effect.

[0096] Furthermore, the control instruction sent by the intelligent analysis server is to directly send a switching signal of the braking system to achieve a braking effect.

[0097] Furthermore, the network transmission device may be a wireless communication module or a wired communication line or both.

[0098] S106, comparing the working lane separation distance measurement value with the initial distance measurement value, and generating a braking instruction when the difference exceeds a preset threshold;

[0099] If the system detects that the container and the truck are not separated, the system will issue a stop electronic control command and sound an alarm.

[0100] Furthermore, the multi-sensor data fusion performed by the intelligent analysis server specifically includes the following steps:

[0101] (1) Calibrate the measurement data of each sensor.

[0102] (2) Filter the calibrated sensor data separately to remove noise.

[0103] (3) Weighting the filtered data. Different sensors are given different weights to indicate their importance in terms of credibility.

[0104] (4) Integrate the weighted sensor data and seek the best estimate to reduce the error.

[0105] (5) Use appropriate fusion algorithm for fusion.

[0106] There are various artificial intelligence algorithms and statistical methods to choose from, such as factor analysis, expectation maximization method EM-Algo, artificial neural network algorithm ANN, etc.

[0107] This application uses lane conditions as latent factors and implements data fusion based on factor analysis. The specific method is as follows: the observed variable is the ranging data collected by the laser scanner at multiple time points, and it is assumed that it is driven by two latent factors:

[0108] Potential factors:

[0109] : The actual distance to the working lane (target lane).

[0110] : interference of adjacent overtaking lane pair to ranging.

[0111] Observation variables:

[0112] : ranging value of laser scanner at the time point .

[0113] Establishing factor model:

[0114]

[0115] Further, the laser scanner is installed in the middle of the rail-mounted crane cantilever, and the laser scanner performs horizontal scanning on the truck lane direction.

[0116] The high-definition monitoring camera is installed in the middle of the rail-mounted crane cantilever, and the information of the vehicles under the cantilever is collected.

[0117] During the unloading operation, when the spreader is in the box, the system laser scanner starts data collection, and at this time, the initial data before lifting is established.

[0118] When the lock is closed and the container is lifted from the truck, the laser scanning range finder detects the gap between the container and the truck pallet in real time, and collects real-time data. At the same time, the monitoring camera installed on the cantilever collects the real-time state of each lane.

[0119] The real-time data collected by the laser scanner and the real-time state of each lane collected by the camera are used as input conditions for data fusion, the confidence of the real-time data of the laser scanner is updated, and the corrected data after multi-sensor fusion is obtained. Through the comparison between the corrected data and the initial data, the judgment operation of whether the container and the truck pallet are separated is performed. If it is detected that the container and the truck are not separated, the system immediately sends a stop command to stop the spreader lifting. When the system controls the spreader to stop rising, the driver can only operate the spreader to descend, and cannot lift. After the spreader is lowered to the loose rope and the box, the lifting restriction is removed, and the system returns to the initial state to start a new round of spreader anti-lifting detection, allowing the staff to open the lock head and continue working.

[0120] When the system based on the multi-sensor data fusion method is working, the steps are as follows:

[0121] Step 1: Start the controller on the rail-mounted crane and each sensor device;

[0122] Step 2: Establish a data fusion model (such as factor analysis method, expectation maximization method EM-Algo, artificial neural network algorithm ANN, etc.) for training and deployment on the intelligent analysis server;

[0123] Third step: the intelligent analysis server continuously receives the data information of the field sensor and processes the received data by algorithm to obtain the detection result. When the server detects a danger, it sends a corresponding alarm signal to the PLC;

[0124] Fourth step: the PLC sends an emergency braking shutdown instruction to the field device to stop the lifting of the spreader;

[0125] Fifth step: the intelligent analysis server simultaneously feeds back the device information to the monitoring system platform terminal.

[0126] S107, transmit the brake instruction to the rail crane control system, and execute the lifting interruption action of the spreader.

[0127] The truck anti-lifting system based on the multi-sensor data fusion technology has the following advantages:

[0128] (1) A variety of sensors are used in cooperation to collect information about the contact state of the container and the drag plate as comprehensively as possible.

[0129] (2) The collected information is processed and comprehensively processed by using an intelligent fusion algorithm.

[0130] (3) Intelligent judgment rules are established according to the regular understanding obtained under specific conditions (i.e. what information is detected under what conditions to determine that the container and the drag plate are in contact.

[0131] (4) Through multi-sensor data intelligent fusion, the problem of false judgment of the lifting of the spreader due to the passing of the truck in the other lane during the overtaking operation in the adjacent lane is solved.

[0132] The application also provides a truck anti-lifting control device based on lane condition factor analysis, comprising:

[0133] The acquisition module collects container ranging data at consecutive time points during the lifting of the spreader by a laser scanner, and synchronously captures multi-lane vehicle information by a camera to generate multi-source input data;

[0134] The model module constructs a double-factor load model according to the multi-source input data, wherein a first factor is a real ranging component of the working lane, and a second factor is an interference component of the adjacent overtaking lane;

[0135] The decomposition module decomposes the covariance matrix of the double-factor load model into a load matrix and an independent noise matrix;

[0136] The rotation module implements orthogonal rotation on the load matrix to separate the load components of the working lane and the adjacent overtaking lane interference;

[0137] a calculation module, configured to calculate a working lane separation ranging value based on the rotated load matrix and the noise matrix;

[0138] a comparison module, configured to compare the working lane separation ranging value with an initial ranging value, and generate a braking instruction when a difference exceeds a preset threshold;

[0139] an execution module, configured to transmit the braking instruction to a track crane control system and execute a spreader hoisting interruption action.

[0140] Further, the acquisition module acquires container ranging data at continuous time points during a spreader container loading stage by a laser scanner and synchronously captures multi-lane vehicle information by a camera to generate multi-source input data, including:

[0141] The laser scanner is installed in a middle region of a track crane trolley and a scanning direction is parallel to a truck lane.

[0142] The camera is installed in a middle region of a track crane jib and covers a working lane below the jib.

[0143] Further, the execution module transmits the braking instruction to the track crane control system and executes the spreader hoisting interruption action, including:

[0144] The braking instruction is sent to the track crane control system by a wired or wireless transmission mode and is synchronously transmitted to a remote monitoring terminal.

[0145] Further, after the calculation module calculates the working lane separation ranging value based on the rotated load matrix and the noise matrix, including:

[0146] When it is detected that the lane interference component load value exceeds a second preset threshold, the number of time points of the multi-source input data is expanded and the working lane separation ranging value is recalculated.

[0147] Further, before the acquisition module acquires container ranging data at continuous time points during a spreader container loading stage by a laser scanner and synchronously captures multi-lane vehicle information by a camera to generate multi-source input data, including:

[0148] When the track crane controller is started, the double-factor load model is deployed to an intelligent analysis server and real-time communication is established with the laser scanner and the camera.

[0149] The foregoing description of the embodiments has been presented for the purpose of illustration and description. It is not intended to be exhaustive or to limit the embodiments to the precise form disclosed. Modifications and variations are possible in light of the above teachings or can be acquired from practice of the embodiments. The embodiments were chosen and described in order to explain the principles of the technology and its practical application and to enable others skilled in the art to understand its application and practice the technology, without limitations as to its details.

Claims

1. A method for preventing container trucks from lifting based on lane condition factor analysis, characterized in that: include: A laser scanner collects container distance data at consecutive time points during the spreader loading phase, and a camera synchronously captures vehicle information on multiple lanes to generate multi-source input data. A two-factor load model is constructed based on the multi-source input data, wherein the first factor is the actual distance measurement component of the working lane, and the second factor is the interference component of the adjacent overtaking lane; Decomposing the covariance matrix of the two-factor loading model into a loading matrix and an independent noise matrix; Performing orthogonal rotation on the load matrix to separate the load components of the working lane and the adjacent passing lane; Calculating a lane separation ranging value based on the rotated load matrix and the noise matrix; Comparing the working lane separation distance value with the initial distance value, and generating a braking command when the difference exceeds a preset threshold; The braking instruction is transmitted to the rail crane control system to execute the lifting interruption action of the spreader.

2. The method for preventing container trucks from lifting based on lane condition factor analysis according to claim 1 is characterized in that: A laser scanner collects container distance data at consecutive time points during the spreader loading phase, and a camera synchronously captures vehicle information on multiple lanes to generate multi-source input data, including: The laser scanner is installed in the middle area of ​​the rail crane trolley, and the scanning direction is parallel to the truck lane; The camera is installed in the middle of the rail crane cantilever, covering the working lane below the cantilever.

3. The method for preventing container trucks from lifting based on lane condition factor analysis according to claim 1 is characterized in that: Transmitting the braking command to the rail crane control system to execute the hoisting interruption action of the spreader includes: The braking command is sent to the rail crane control system via wired or wireless transmission, and is synchronously transmitted to the remote monitoring terminal.

4. The method for preventing container trucks from lifting based on lane condition factor analysis according to claim 1 is characterized in that: The laser scanner collects container distance measurement data at consecutive time points during the spreader loading phase, and the camera synchronously captures vehicle information in multiple lanes. Before generating multi-source input data, the following are included: When the track crane controller is started, the dual-factor load model is deployed to the intelligent analysis server, and real-time communication is established with the laser scanner and the camera.

5. A container truck anti-lifting control device based on lane condition factor analysis, characterized in that: include: The acquisition module uses a laser scanner to collect container distance measurement data at consecutive time points during the spreader loading phase, and uses a camera to synchronously capture multi-lane vehicle information to generate multi-source input data; A model module constructs a dual-factor load model based on the multi-source input data, wherein the first factor is a real distance measurement component of the working lane, and the second factor is an interference component of the adjacent overtaking lane; A decomposition module decomposes the covariance matrix of the two-factor load model into a load matrix and an independent noise matrix; A rotation module performs orthogonal rotation on the load matrix to separate the load components of the working lane and the adjacent passing lane; a calculation module, calculating a working lane separation ranging value based on the rotated load matrix and the noise matrix; a comparison module, for comparing the working lane separation distance value with the initial distance value, and generating a braking instruction when the difference exceeds a preset threshold; The execution module transmits the braking instruction to the rail crane control system to execute the lifting interruption action of the spreader.

6. The anti-lifting control device for container trucks based on lane condition factor analysis according to claim 5 is characterized in that: The acquisition module uses a laser scanner to collect container distance measurement data at consecutive time points during the spreader loading phase, and uses a camera to synchronously capture multi-lane vehicle information to generate multi-source input data, including: The laser scanner is installed in the middle area of ​​the rail crane trolley, and the scanning direction is parallel to the truck lane; The camera is installed in the middle of the rail crane cantilever, covering the working lane below the cantilever.

7. The anti-lifting control device for container trucks based on lane condition factor analysis according to claim 5 is characterized in that: The execution module transmits the braking instruction to the rail crane control system to execute the hoisting interruption action of the spreader, including: The braking command is sent to the rail crane control system via wired or wireless transmission, and is synchronously transmitted to the remote monitoring terminal.

8. The anti-lifting control device for container trucks based on lane condition factor analysis according to claim 5 is characterized in that: The acquisition module collects container distance measurement data at consecutive time points during the spreader loading phase using a laser scanner, and synchronously captures multi-lane vehicle information using a camera to generate multi-source input data, including: When the track crane controller is started, the dual-factor load model is deployed to the intelligent analysis server, and real-time communication is established with the laser scanner and the camera.

Images