Link pre-switching method and device based on spatial position

Abstract

The invention discloses a link pre-switching method and device based on a spatial position, relates to the technical field of port equipment communication, and mainly aims to meet the requirements of low time delay and high reliability of remote control of a tire crane. According to the technical scheme, the method comprises the steps of obtaining multi-source operation data of a target tire crane; on the basis of the multi-source operation data, a predicted space-time operation state corresponding to the target tire crane in a specified time period in the future is pre-judged through a space-time joint model, and the predicted space-time operation state comprises a predicted space position, a predicted attitude parameter and a predicted shielding risk probability; calculating a link switching decision score of the target tire crane through a multi-dimensional decision matrix based on the predicted space-time operation state; if the link switching decision score reaches a preset score threshold, triggering a pre-switching operation; and switching a communication link of the target tire crane from the original base station to the target base station after the target tire crane reaches the predicted spatial position and the signal strength and the transmission delay of the target base station are verified to meet preset transmission requirements.

Description

Technical Field

[0001] This application relates to the field of port equipment communication technology, and in particular to a link pre-switching method and apparatus based on spatial location. Background Technology

[0002] With the accelerated transformation of global ports towards automation and intelligence, remote control technology for rubber-tired gantry cranes (LTGs), as core equipment for container handling, has become crucial for improving port operational efficiency and reducing labor costs. In remote control scenarios, wireless communication between the LTG and the control room must meet the core requirements of high bandwidth, low latency, and high reliability. Millimeter-wave communication, due to its advantages of high spectrum resources, large bandwidth, and low latency, is gradually becoming the mainstream choice for port wireless communication.

[0003] Currently, existing technologies for wireless link switching in port rubber-tired gantry cranes mostly employ reactive switching. This involves real-time monitoring of the signal strength of the current base station (original base station), triggering a link switch to another available base station (target base station) when the signal strength falls below a preset threshold. However, this reactive switching method does not incorporate predictive information such as the gantry crane's movement trajectory and operating posture, lacking a pre-preparation phase. Consequently, link switching relies on real-time signal attenuation triggering, resulting in high switching latency and susceptibility to signal interruption due to the narrow beam characteristics of millimeter waves. This makes it difficult to meet the low latency and high reliability requirements of remote control of rubber-tired gantry cranes. Summary of the Invention

[0004] In view of the above problems, this application provides a link pre-switching method and apparatus based on spatial location. The main purpose is to reduce link switching delay, avoid signal interruption, and meet the low latency and high reliability requirements of remote control of tire cranes.

[0005] To solve the above-mentioned technical problems, this application proposes the following solution: Firstly, this application provides a spatial location-based link pre-switching method applied to a port tire crane wireless communication system, the method comprising: Acquire multi-source operational data of the target tire crane; Based on the multi-source operation data, the spatiotemporal operating status of the target tire crane within a specified future time period is predicted by a spatiotemporal joint model. The predicted spatiotemporal operating status includes the predicted spatial location, predicted attitude parameters, and predicted occlusion risk probability. Based on the predicted spatiotemporal operating status, the link switching decision score of the target tire crane is calculated through a multidimensional decision matrix. The decision factors of the multidimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical switching success rate and weather attenuation compensation coefficient. If the link handover decision score reaches a preset score threshold, a pre-handover operation is triggered. The pre-handover operation includes at least selecting a corresponding target base station in the base station group based on the predicted spatial location and the historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location. Once the target tire crane reaches the predicted spatial location and verifies that the signal strength and transmission delay of the target base station meet the preset transmission requirements, the communication link of the target tire crane is switched from the original base station to the target base station.

[0006] Secondly, this application provides a spatial location-based link pre-switching device for use in a port rubber-tired gantry crane wireless communication system, the device comprising: The acquisition unit is used to acquire multi-source operation data of the target tire crane; The prediction unit is used to predict the predicted spatiotemporal operating state of the target tire crane within a specified future time period based on the multi-source operation data obtained by the acquisition unit and through a spatiotemporal joint model. The predicted spatiotemporal operating state includes the predicted spatial location, predicted attitude parameters, and predicted occlusion risk probability. The calculation unit is used to calculate the link switching decision score of the target tire crane based on the predicted spatiotemporal operating status obtained by the prediction unit through a multi-dimensional decision matrix. The decision factors of the multi-dimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical switching success rate and weather attenuation compensation coefficient. A triggering unit is used to trigger a pre-handover operation if the link handover decision score obtained by the calculation unit reaches a preset score threshold. The pre-handover operation includes at least selecting a corresponding target base station in the base station group based on the predicted spatial location and the historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location. The first control unit is used to switch the communication link of the target tire crane from the original base station to the target base station obtained by the triggering unit after the target tire crane reaches the predicted spatial position and verifies that the signal strength and transmission delay of the target base station meet the preset transmission requirements.

[0007] To achieve the above objectives, according to a third aspect of this application, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to perform the spatial location-based link pre-switching method of the first aspect described above.

[0008] To achieve the above objectives, according to a fourth aspect of this application, a processor is provided for running a program, wherein the program executes the spatial location-based link pre-switching method of the first aspect described above.

[0009] Using the above technical solution, this application provides a link pre-switching method and apparatus based on spatial location. First, multi-source operational data of the target rubber-tired crane is acquired. Then, a spatiotemporal joint model is used to predict the predicted spatiotemporal operating state of the target rubber-tired crane within a specified future time period. The predicted spatiotemporal operating state includes the predicted spatial location, predicted attitude parameters, and predicted obstruction risk probability. Next, based on the predicted spatiotemporal operating state, a multi-dimensional decision matrix is ​​used to calculate the link switching decision score of the target rubber-tired crane. The decision factors of the multi-dimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical handover success rate, and weather attenuation compensation coefficient. If the link switching decision score reaches a preset score threshold, a pre-switching operation is triggered. The pre-switching operation includes at least selecting a corresponding target base station from the base station group based on the predicted spatial location and historical handover success rate, and controlling the beam direction of the target base station to pre-align and cover the predicted spatial location. Finally, when the target rubber-tired crane arrives at the predicted spatial location and the signal strength and transmission delay of the target base station are verified to meet preset transmission requirements, the communication link of the target rubber-tired crane is switched from the original base station to the target base station. The technical solution provided in this application acquires multi-source operational information of the rubber-tired gantry crane to provide comprehensive data support for accurate prediction. By using a spatiotemporal joint model, it predicts the spatial location, attitude parameters, and obstruction risk probability of the rubber-tired gantry crane in a specified future time period, realizing predictive handover. Based on a multi-dimensional decision matrix containing multiple decision factors, it calculates the link handover decision score, effectively improving the accuracy of link decision. After triggering the pre-handover, it selects the target base station based on the predicted spatial location and historical handover success rate and pre-aligns the beam in advance, eliminating the time spent on beam adjustment during handover and significantly reducing handover delay. After the rubber-tired gantry crane arrives, it verifies the signal strength and transmission delay before completing the handover, further ensuring handover reliability. Ultimately, it effectively solves the problems of high handover delay and easy signal interruption due to millimeter wave obstruction in the existing technology, effectively reducing handover delay and avoiding signal interruption, meeting the low latency and high reliability requirements of remote control of port rubber-tired gantry cranes.

[0010] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1A flowchart of a link pre-switching method based on spatial location provided in an embodiment of this application is shown; Figure 2 This paper illustrates a flowchart of another spatial location-based link pre-switching method provided in an embodiment of this application. Figure 3 This illustration shows a block diagram of a spatial location-based link pre-switching device according to an embodiment of this application; Figure 4 This paper illustrates a block diagram of another spatial location-based link pre-switching device provided in an embodiment of this application. Detailed Implementation

[0012] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0013] Currently, existing technologies for wireless link switching in port rubber-tired gantry cranes mostly employ reactive switching. This involves real-time monitoring of the signal strength of the current base station (original base station), triggering a link switch to another available base station (target base station) when the signal strength falls below a preset threshold. However, this reactive switching method does not incorporate predictive information such as the gantry crane's movement trajectory and operating posture, lacking a pre-preparation phase. Consequently, link switching relies on real-time signal attenuation triggering, resulting in high switching latency and susceptibility to signal interruption due to the narrow beam characteristics of millimeter waves. This makes it difficult to meet the low latency and high reliability requirements of remote control of rubber-tired gantry cranes.

[0014] Research has shown that by using multi-source operational data to drive spatiotemporal state prediction and multi-dimensional decision-driven pre-switching, target base station selection and beam pre-alignment can be completed in advance. After the tire crane arrives, signal switching can be verified, thereby realizing the transformation from reactive to predictive switching. This significantly reduces switching latency, avoids signal interruption, and meets the low latency and high reliability requirements of remote control of tire cranes.

[0015] Based on the above considerations, this application provides a link pre-switching method based on spatial location. This method can reduce link switching latency and avoid signal interruption, meeting the low latency and high reliability requirements of remote control of rubber-tired cranes. It is applied to a port rubber-tired crane wireless communication system. The core architecture of the system includes: a data acquisition unit at the rubber-tired crane end (including a PLC controller, a UWB positioning module, and a lidar), a base station group (12 60GHz millimeter-wave base stations + 3 5.8GHz hot standby base stations), and a central control unit (including a GiNMS operation and maintenance platform, a spatiotemporal joint model inference module, and a link switching scheduling unit). The specific execution steps are as follows: Figure 1As shown, it includes: 101. Obtain multi-source operation data of the target tire crane.

[0016] The multi-source operational data includes PLC control signals, UWB positioning data, and path planning data from the dock scheduling system.

[0017] In this step, for PLC control signals, the steering angle, travel speed, and gantry lifting height are collected in real time via the PROFINET interface of the PLC controller integrated with the tire crane, at a frequency of 100Hz. The data is then transmitted to the central control unit via Ethernet. For UWB positioning data, UWB positioning tags are installed on the top of the tire crane gantry, and eight UWB anchor points are deployed around the yard. The TDOA positioning algorithm is used to calculate the current spatial position (X / Y coordinates in the local coordinate system of the yard, in meters) and position error (range 0-8cm) of the tire crane in real time. The positioning data is then synchronously uploaded via a 5.8GHz backup link. For the path planning data of the terminal scheduling system, the preset operation path (such as the coordinate sequence of “Block2-5 -> Block3-3 -> Block4--1”), docking nodes (such as (85.6, 42.3), (102.4, 56.8)) and operation duration (such as 120s for relocation and 80s for hoisting) of the target rubber-tired gantry crane are obtained through the API interface of the terminal TOS scheduling system. The data update frequency is 1Hz, and real-time push is triggered when the scheduling path is adjusted.

[0018] After obtaining the PLC control signals, UWB positioning data, and dock scheduling system path planning data of the target rubber-tired gantry crane, preprocessing can be performed, including format standardization, time synchronization, and noise reduction and enhancement. Specifically, the analog quantities of the PLC control signals (e.g., speed 12.5 km / h) are normalized to the 0-1 range, the UWB latitude and longitude are converted to the local coordinate system of the yard, and the scheduling path text is converted to a coordinate array. Using the timestamp of the UWB positioning data as a reference, the time nodes of the PLC control signals and the scheduling system path planning data are completed through linear interpolation. Kalman filtering is used to filter out vibration noise in the PLC speed signals, and outliers in the UWB positioning data are removed using the 3σ criterion (e.g., values ​​with an error > 10 cm are directly replaced with the average of the first 3 nodes). This yields a multi-source operation data sequence, such as speed changes, position trajectories, and scheduling path nodes within the past 200 ms.

[0019] 102. Based on multi-source operation data, predict the spatiotemporal operating status of the target tire crane within a specified future time period using a spatiotemporal joint model.

[0020] Among them, predicting the spatiotemporal operating state includes predicting the spatial location, predicting the attitude parameters, and predicting the probability of occlusion risk.

[0021] In this embodiment, the spatiotemporal joint model is a trained bidirectional LSTM network deployed in the central control unit. The network parameters are as follows: the input layer dimension is a multi-source job data sequence of the past 200ms (40 time steps × 6 feature dimensions, including velocity, position, turning angle, etc.), the hidden layer is a 2-layer 128-dimensional neuron with a dropout rate of 0.2, the output layer is a 3-dimensional vector (predicting spatial position X / Y, predicting pose parameters, and predicting occlusion risk probability), and the inference time is ≤10ms.

[0022] In this step, the multi-source operation data sequence preprocessed in step 101 (such as speed changes, position trajectories, and scheduling path nodes within the past 200ms) is used as input data. The spatiotemporal joint model first performs weighted fusion of PLC control signals, UWB positioning data, and terminal scheduling system path planning data through a temporal attention mechanism. Then, it extracts the temporal correlation features of "position-velocity-attitude" through a bidirectional LSTM network, and finally outputs the prediction results for a specified future time period, including predicted spatial position, predicted attitude parameters, and predicted occlusion risk probability. The predicted spatial position is the X / Y coordinate accurate to 0.1m, with a position prediction error ≤10cm. The predicted attitude parameters are the steering angle and gantry height, with an attitude error ≤5%. The predicted occlusion risk probability is based on the obstacle data of the yard scanned by LiDAR (such as container stacking height of 7.2m), outputting a probability value of 0-100% (e.g., 75%), with a prediction accuracy ≥95%.

[0023] 103. Based on the predicted spatiotemporal operating status, calculate the link switching decision score of the target tire crane through a multi-dimensional decision matrix.

[0024] The decision factors of the multidimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical handover success rate, and weather attenuation compensation coefficient.

[0025] In this embodiment, the signal attenuation slope refers to the rate of change within the last 50ms calculated based on the real-time RSSI data of the base station, such as (-53.5-(-52)) / 50=-0.03dB / ms, and the absolute value is taken as the quantization value; the occlusion risk index refers to the predicted occlusion risk probability output in step 102 (such as 75%); the service priority refers to the quantization based on the "proportion of key instructions" in the PLC control signal, such as the proportion of lifting / braking instructions 60% -> quantization value 1.0, the proportion of video stream 50% -> 0.7, and the proportion of status data 70% -> 0.3; the historical handover success rate refers to the handover data of the target base station in the last 3 months retrieved from the GiNMS platform under the "predicted location ±5m, same occlusion scenario", such as 1198 successful handovers out of 1200 -> success rate 99.83%; the weather attenuation compensation coefficient refers to the data of the associated wharf meteorological station, such as light rain -> 1.2, moderate rain -> 1.5, heavy rain -> 2.0, and sunny -> 1.0.

[0026] The quantified values ​​are converted into scores of 0-100 according to the preset single-factor scoring rules, resulting in a single-factor score for each decision factor. For example, signal attenuation slope 0.03dB / ms → 99 points, obstruction risk index 75% → 90 points, business priority 1.0 → 100 points, historical success rate 99.83% → 99.83 points, and weather coefficient 1.2 → 90 points. A pre-set weight allocation rule automatically adjusts the weight percentage based on the current port scenario. For example, in this scenario of "medium-speed transfer + light rain + hoisting operation," the weight allocation would be: signal attenuation slope 30%, obstruction risk index 25%, business priority 20%, historical handover success rate 15%, and weather attenuation compensation coefficient 10%.

[0027] In this step, the single-factor scores of each decision factor and the weight percentage of each decision factor obtained from the weight allocation rules are weighted and summed to obtain the link switching decision score. This link switching decision score is used to characterize the comprehensive degree of necessity and feasibility of the target tire crane switching the communication link at the preset spatial location. The higher the score, the stronger the necessity and the higher the feasibility of switching.

[0028] 104. If the link switching decision score reaches the preset score threshold, a pre-switching operation is triggered.

[0029] The pre-handover operation includes at least selecting the corresponding target base station in the base station group based on the predicted spatial location and historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location.

[0030] In this step, the preset score threshold is a critical standard for quantifying the necessity and feasibility of link handover, used to determine whether to trigger a pre-handover operation. The preset score threshold can be 70, 80, etc., and can be verified through multiple field tests or customized based on usage requirements. It is only necessary to ensure that this threshold balances handover sensitivity and false handover rate. When the link handover decision score is greater than or equal to the preset score threshold, it indicates that the necessity and feasibility of handover to the target base station are strong, meeting the requirements for low-latency and high-reliability communication. Therefore, pre-handover is triggered, and the central control unit sends a pre-handover command to the base station group via the TCP / IP protocol.

[0031] The base station cluster consists of 12 60GHz millimeter-wave base stations (deployed atop the yard lighthouse, each with a coverage radius of 150m and supporting phase-reconfigurable array antennas) and 3 5.8GHz hot standby base stations (covering the entire yard as redundant backups). Based on the predicted spatial location, at least one millimeter-wave base station covering that location is selected as a candidate base station. The historical handover success rate of the candidate base stations is retrieved, and the candidate base station with the highest historical handover success rate is selected as the target base station. Based on the installation coordinates of the target base station and the predicted spatial location, beam pointing parameters, including horizontal angle, elevation angle, and beamwidth, are calculated using spatial geometric formulas. The central control unit issues beam adjustment commands to the target base station, and the Xilinx Kintex-7 FPGA processing unit built into the base station adjusts the phase of 16 antenna elements within 3.1μs to complete beam pre-alignment. Laser ranging verification shows that the error in beam coverage of the predicted spatial location is 8cm≤10cm, and the pre-alignment completion time is 120ms ahead of the arrival of the tire crane at the predicted location.

[0032] 105. Once the target tire crane reaches the predicted spatial location and the signal strength and transmission delay of the target base station are verified to meet the preset transmission requirements, the communication link of the target tire crane is switched from the original base station to the target base station.

[0033] In this step, when UWB positioning data shows that the target tire crane has reached the predicted spatial location, signal verification can be triggered. Specifically, the millimeter-wave remote device of the target tire crane collects signal indicators from the target base station in real time. It can collect 10 consecutive sets of signal strength (RSSI) and transmission delay, calculate their average, and set a first preset threshold for signal strength and a second preset threshold for transmission delay as preset transmission requirements. Verification passes when the signal strength is greater than or equal to the first preset threshold and the transmission delay is less than or equal to the second preset threshold; otherwise, verification fails. It should be noted that when verification fails, the target reference beam can be fine-tuned and a second verification performed. Alternatively, a new candidate base station can be selected as the new target base station, and pre-alignment and signal verification can be repeated. A 5.8GHz hot standby base station can also be activated, and a handover or fallback to the original base station can be completed, triggering an alarm. This embodiment does not limit these possibilities.

[0034] A switch is used as the central redundant switching unit. Service transmission at the original base station is suspended, the link to the target base station is simultaneously activated, and data is migrated according to the priority order of "control command -> video stream -> status data" to complete the handover from the original base station to the target base station. After the handover, the GiNMS platform continuously monitors the signal indicators of the target base station. The original base station can remain in standby mode for a preset time period. If the signal indicators of the target base station continuously meet the preset transmission requirements within the preset time period, the target base station will deactivate its standby mode. If not, the 5.8GHz hot standby base station can be activated and the handover completed, or the system can fall back to the original base station and trigger an alarm.

[0035] Based on the above Figure 1 As can be seen from the implementation method, the link pre-switching method based on spatial location provided in this application obtains multi-source operation information such as PLC control signals, UWB positioning data, and scheduling path planning data to provide comprehensive data support for accurate prediction. It also uses a spatiotemporal joint model to predict the spatial location, attitude parameters, and obstruction risk probability of the rubber-tired gantry crane in a specified period of time in the future, realizing predictive switching. The link switching decision score is calculated based on a multi-dimensional decision matrix containing multiple decision factors, which effectively improves the accuracy of link decision. After triggering the pre-switching, the target base station is selected based on the predicted spatial location and historical switching success rate, and the beam is pre-aligned in advance, saving the time spent on beam adjustment during switching and greatly reducing the switching delay. After the rubber-tired gantry crane arrives, the signal strength and transmission delay are verified before the switching is completed, which further ensures the reliability of switching. Finally, it effectively solves the problems of high switching delay and easy signal interruption due to millimeter wave obstruction in the prior art, effectively reduces switching delay, avoids signal interruption, and meets the low latency and high reliability requirements of remote control of rubber-tired gantry cranes in ports.

[0036] Furthermore, the preferred embodiments of this application are based on the above... Figure 1 Based on this, a detailed explanation of the link pre-handover process based on spatial location is provided, and the specific steps are as follows: Figure 2 As shown, it includes: 201. Obtain multi-source operation data of the target tire crane.

[0037] This step combines the description of step 101 in the above method, and the same content will not be repeated here.

[0038] 202. Based on multi-source operation data, predict the spatiotemporal operating status of the target tire crane within a specified future time period using a spatiotemporal joint model.

[0039] This step combines the description of step 102 in the above method, and the same content will not be repeated here.

[0040] Furthermore, before predicting the spatiotemporal operating state of the target rubber-tired crane in a future period using a spatiotemporal joint model based on multi-source operation data, the process includes: preprocessing the multi-source operation data and adding labels corresponding to the actual spatiotemporal operating state to the preprocessed multi-source operation data within a specified historical period. The actual spatiotemporal operating state includes the actual spatial location, actual attitude parameters, and actual occlusion risk probability. The labeled multi-source operation data is then weighted and fused, and the corresponding temporal features are extracted to obtain temporal data. The temporal data is used as samples to construct and train a bidirectional LSTM network corresponding to the target rubber-tired crane to obtain a spatiotemporal joint model. This spatiotemporal joint model is used to output the predicted spatiotemporal operating state of the target rubber-tired crane in a specified future period.

[0041] In this step, differentiated processing strategies are adopted for preprocessing based on the different characteristics of PLC control signals, UWB positioning data, and dock scheduling path planning data. Specifically, for PLC control signal preprocessing: the raw data contains mechanical vibration noise (such as speed fluctuations of ±0.3km / h during gantry lifting) and electromagnetic interference spikes (occasional jumps in steering angle of ±5°). The processing flow is as follows: Kalman filtering is used for noise reduction (state equation: X(k)=AX(k-1)+Bu(k)+w(k), observation equation: Z(k)=H*X(k)+v(k), filter parameters Q=0.01, R=0.1) to filter out high-frequency noise in speed and steering angle; linear interpolation is used to fill in missing values ​​(1-2 data points missing due to instantaneous communication interruption); normalization processing: the steering angle, moving speed, and gantry height are mapped to the [0,1] interval, and the formula is: normalized value = (original value - minimum value) / (maximum value - minimum value).

[0042] For UWB positioning data preprocessing: The raw data contains positioning drift (such as occasional position errors of 15-20cm) and outliers caused by anchor point signal obstruction. The processing flow is as follows: outliers are removed using the 3σ criterion (the mean position error μ=5cm and the standard deviation σ=2cm are calculated, and data points with errors >μ+3σ=11cm are removed); the moving average filter (window size of 5 data points) is used to smooth the position trajectory and reduce positioning jitter; coordinate transformation: the original UWB latitude and longitude (such as 28.12°N, 121.56°E) are converted to the local coordinate system of the storage yard (X / Y axis unit m) through Gauss-Kruger projection, with a transformation error ≤2cm.

[0043] For the preprocessing of terminal scheduling route planning data: the original data is in text format (e.g., "Block2-5—>Block3-3, operation time 120s"). The processing flow is as follows: extract the coordinates of the docking nodes (e.g., "Block2-5" corresponds to the predefined coordinates (62.3, 35.8)), operation time, and path priority through regular expressions; convert the discrete path nodes into a coordinate sequence according to the operation time (e.g., [(62.3, 35.8), (70.1, 38.2), ..., (85.6, 42.3)]), and complete the intermediate nodes to form continuous path data.

[0044] Using the timestamps of UWB positioning data as the reference clock, multi-source data synchronization is achieved through "interpolation completion + timestamp alignment". For example, for PLC control signals, linear interpolation is used to complete missing data points; for dock scheduling path planning data, constant padding is used to expand it into interval data (keeping the coordinates of the previous node unchanged when the work path remains the same). After synchronization, all data is uniformly marked with a timestamp, with the format: YYYY-MM-DDHH:MM:SS:SSS.

[0045] For the preprocessed multi-source operational data, tagged samples are constructed. The tags represent the actual spatiotemporal operating status of the tire-mounted crane within a specified historical time period. These tags are obtained through high-precision measured data, ensuring accurate alignment with the preprocessed multi-source operational data. This results in a sufficient number of effective samples covering all scenarios, including sunny / rainy days, high / low speeds, and high / low occlusion. Specifically, these samples include actual spatial location, actual attitude parameters, and actual occlusion risk probability. For the actual spatial location acquisition, UWB positioning data is preprocessed and combined with distance measurement data from LiDAR at key nodes of the tire-mounted crane (such as the top of the gantry), using triangulation to correct position errors. For the actual attitude parameters, historical execution data (actual execution values ​​of steering angle and gantry height) from the tire-mounted crane PLC controller is directly read and compared with the attitude sensor data installed in the tire-mounted crane cab. Outliers are removed, and the remaining data are used as the actual attitude tags. To collect the actual probability of occlusion risk, the surrounding environment of the rubber-tired crane is scanned in real time by lidar, and the distance and height between the obstruction (such as containers or trucks) and the rubber-tired crane are calculated. Combined with the actual attenuation value of the millimeter-wave base station signal, the agile attenuation value specifically refers to the difference between the measured RSSI value and the value when there is no obstruction, and a logistic regression model is used to calculate the actual probability of occlusion risk.

[0046] Samples can be constructed in 200ms time windows. Each sample contains preprocessed multi-source data from the past 200ms (40 time steps × 6 features: velocity, steering angle, gantry height, X coordinate, Y coordinate, path priority) and the actual spatiotemporal running state for the next 500ms (labels: X / Y coordinates of the destination in the next 500ms, pose parameters, occlusion risk probability), ensuring the temporal correlation between the input and the label. The dataset is divided into training, validation, and test sets in a 7:2:1 ratio. The training set covers the entire scene, the validation set is used to adjust model hyperparameters, and the test set is used for final performance evaluation. By weighted fusion integrating complementary information from multi-source data and extracting temporal features to capture the dynamic correlation between "position-velocity-pose," high-quality input is provided to the bidirectional LSTM network. Weights are assigned based on the contribution of each data point to the prediction target (calculated using mutual information entropy: UWB positioning data has the highest mutual information entropy with location prediction, weight 0.35; PLC control signals are second, weight 0.4; scheduling data weight 0.25). A weighted average method is used to fuse features: fused feature value = 0.4 × PLC feature value + 0.35 × UWB feature value + 0.25 × scheduling feature value. Features are extracted using a combination of sliding window extraction and time-series statistics to meet the time-series modeling requirements of bidirectional LSTM networks: Sliding window settings: 20 time steps per window, with a sliding step size of 5 time steps, extracting trend features (such as velocity change rate, position offset) and statistical features (such as mean velocity, steering angle variance) within each window; Feature dimensionality reduction: Principal component analysis reduces high-dimensional time-series features from 24 dimensions to 12 dimensions, retaining 98% of the feature information and reducing model computation; the final output time-series data format is "number of samples × number of time steps × number of features".

[0047] A bidirectional LSTM network was built based on the TensorFlow 2.10 framework. Network parameters were optimized through training to enable the model to accurately learn the mapping relationship between multi-source temporal data and future spatiotemporal states. For temporal data with "40 time steps × 12 features", the input data underwent batch normalization to accelerate model convergence. Two hidden layers were used: the first layer had 128-dimensional neurons, and the second layer had 64-dimensional neurons. A "forward LSTM + backward LSTM" structure was adopted, where the forward LSTM learns the positive temporal correlations of historical data, such as the influence of past velocity on future position, and the backward LSTM learns the negative correlations, such as the constraints of future path nodes on the current pose. The bidirectional outputs were concatenated to enhance feature representation. A dropout layer (with a dropout rate of 0.2) can be added between the two LSTM layers to suppress overfitting and improve the model's generalization ability. The system employs two fully connected layers: the first layer consists of 32-dimensional neurons (using ReLU activation function), and the second layer is a 3-dimensional output layer (using Linear activation function for position, Sigmoid activation function for pose, and Sigmoid activation function for occlusion risk), corresponding to three prediction targets. A multi-task loss function is used to combine the errors of the three prediction targets. For example, the total loss = 0.4 × position prediction MSE loss + 0.3 × pose prediction MAE loss + 0.3 × occlusion risk cross-entropy loss. Among these, the position loss has the highest weight because pre-switching requires the highest accuracy in position prediction. The Adam optimizer is selected with an initial learning rate of 0.001, dynamically adjusted using a cosine annealing strategy to avoid training stagnation. The batch size is set to 64, and training lasts for 50 epochs. An early stopping strategy is used: training stops if the validation set loss does not decrease for 5 consecutive epochs. The training set / validation set loss curves are monitored in real time during training to avoid overfitting. For low-accuracy scenarios, transfer learning is used to supplement the model with samples from that scenario for fine-tuning.

[0048] The spatiotemporal joint model constructed above can accurately predict the position, attitude, and occlusion risk probability of the target tire crane within a specified time period in the future, providing core data support for accurate decision-making on subsequent link pre-switching.

[0049] 203. Obtain the single-factor scoring rules corresponding to each decision factor, and calculate the single-factor score of each decision factor based on the single-factor scoring rules.

[0050] In this step, the single-factor scoring rule quantifies the value of each decision factor among signal attenuation slope, obstruction risk index, service priority, historical handover success rate, and weather attenuation compensation coefficient, converting them into a standard score of 0-100. To ensure that the single-factor score accurately reflects the impact of the decision factor on the handover decision, the specific rules are shown in Table 1: Table 1 Decision factors Quantitative indicator definition Scoring rules (0-100 points) Signal attenuation slope Absolute value of RSSI rate of change over the past 50 ms (dB / ms) ≤0.05 → 100 points; 0.05-0.1 → 90 points; 0.1-0.3 → 70 points; 0.3-0.5 → 40 points; >0.5 → 20 points (the faster the decay, the lower the score) Obstruction Risk Index The predicted occlusion risk probability (%) output by the spatiotemporal joint model ≤50—>100 points; 50-70—>90 points; 70-80—>75 points; 80-90—>40 points; >90—>20 points (the lower the risk of occlusion, the higher the score). Business Priority Percentage of critical instructions in PLC control signals (%) ≥40 (lifting / braking commands) —> 100 points; 20-40 (mainly video stream) —> 80 points; <20 (mainly status data) —> 60 points Historical switch success rate Successful handover percentage within ±5m of the same location and in the same scenario (%) ≥99.5 → 100 points; 99-99.5 → 90 points; 98-99 → 70 points; 95-98 → 40 points; <95 → 20 points (the higher the success rate, the higher the score) Weather attenuation compensation coefficient Millimeter wave attenuation correction factor based on rainfall 1.0 (Sunny) —> 100 points; 1.0-1.5 (Light rain / Cloudy) —> 90 points; 1.5-2.0 (Moderate rain) —> 70 points; >2.0 (Heavy rain) —> 40 points 204. Based on the preset weight allocation rules, the single-factor scores of each decision factor are weighted and summed to obtain the link switching decision score.

[0051] The weighting rules are dynamically changed according to the different port scenarios corresponding to the target rubber-tired gantry crane. The different port scenarios include high-speed transfer scenarios, rainy day operation scenarios, core control operation scenarios, and high obstruction scenarios.

[0052] In this step, the weight allocation rule is to automatically adjust the weight ratio of each factor according to the current operating scenario of the rubber-tired gantry crane in the port, and obtain the link switching decision score by weighted summation. Both weight adjustment and score calculation are deployed in the central control unit.

[0053] Based on four typical scenarios—high-speed transfer scenario, rainy weather operation scenario, core control operation scenario, and high occlusion scenario—a differentiated weighting table was developed, with the total weight always set to 100%. The specific rules are shown in Table 2. Table 2 Scene type Triggering conditions Weighting (signal attenuation: obstruction risk: business priority: historical success rate: weather factor) High-speed transition scene Moving speed ≥ 50km / h 40%:15%:20%:15%:10% Working in rainy weather Rainfall ≥ 30 mm / h 30%:20%:20%:15%:15% Core control scenario Critical instructions account for ≥40% 30%:25%:30%:10%:5% High occlusion scene The probability of occlusion is ≥80%. 25%:35%:20%:20%:0% Default scenario (medium speed + light rain + core control) Not meeting the requirements of a single scenario, but multiple scenarios combined. 30%:25%:20%:15%:10% The dynamic adjustment of weight allocation rules is triggered by preset scene thresholds, including: movement speed threshold, rainfall threshold, critical command percentage threshold, and occlusion risk probability threshold. When the detected scene data reaches the corresponding threshold, the weight configuration for that scene is automatically activated. The specific threshold is not a fixed value and can be customized as needed.

[0054] Furthermore, based on the preset weight allocation rules, the weighted summation of the single-factor scores of each decision factor to obtain the link handover decision score is performed as follows: The initial weight percentages corresponding to different decision factors in the weight allocation rules are obtained; when in a high-speed transition scenario, the initial weight percentage of the signal attenuation slope is increased according to the preset gradient, and the initial weight percentage of the historical handover success rate is decreased according to the preset gradient; when in a rainy operation scenario, the initial weight percentage of the weather attenuation compensation coefficient is increased according to the preset gradient, and the initial weight percentage of the occlusion risk index is decreased according to the preset gradient; when in a core control operation scenario, the weight percentage of the business priority is increased according to the preset gradient, and the initial weight percentages of the signal attenuation slope, occlusion risk index, historical handover success rate, and weather attenuation compensation coefficient are proportionally allocated; when in a high occlusion scenario, the initial weight percentage of the occlusion risk index is increased according to the preset gradient, and the initial weight percentage of the historical handover success rate is increased according to the preset gradient, and the initial weight percentages of the signal attenuation slope, business priority, and weather attenuation compensation coefficient are proportionally allocated.

[0055] In this step, the preset gradient is a fixed adjustment range, which can be verified by actual testing to ensure that the adjusted weights can accurately match the needs of the scenario. The single gradient adjustment range is 5%-15%, and when adjusting multiple factors, the total weight must always be 100%.

[0056] For high-speed relocation scenarios, due to the rapid signal attenuation during high-speed movement, priority should be given to signal attenuation trends, reducing the weight of historical handover success rates. Specifically, the weight of signal attenuation slope can be increased by 15 percentage points, the weight of historical handover success rates reduced by 10 percentage points, and the remaining 5 percentage points distributed among the occlusion risk index and weather attenuation compensation coefficient. For rainy operation scenarios, since rain exacerbates millimeter-wave signal attenuation, the weight of weather factors needs to be increased. Simultaneously, the prediction error of occlusion risk increases in rainy weather, so its weight should be reduced. Specifically, the weight of the weather attenuation compensation coefficient can be increased by 10 percentage points, the weight of the occlusion risk index reduced by 5 percentage points, and the remaining 5 percentage points reduced from historical handover success rates. For core control operation scenarios, since critical commands such as lifting and braking require absolute reliability, the weight of business priority needs to be significantly increased; the weights of other factors should be reduced proportionally to ensure fairness in the adjustment. Specifically, the weight of business priority can be increased by 15 percentage points, and the remaining four decision factors should be reduced by 15 percentage points proportionally to their initial weights. For scenarios with high obstruction, since link stability depends on the base station's anti-obstruction capability and obstruction risk prediction, the weights of both need to be increased simultaneously. Other non-core decision factors should be reduced proportionally. Specifically, the obstruction risk index weight should be increased by 10 percentage points, the historical handover success rate weight by 10 percentage points, and the remaining 20 percentage points should be reduced proportionally from signal attenuation, service priority, and weather coefficient according to their initial weights. Furthermore, it should be noted that when the target tire crane simultaneously meets multiple of the above scenarios, a principle of prioritizing core scenarios supplemented by gradient adjustment can be adopted to ensure that the weights adapt to the needs of the complex scenarios.

[0057] By dynamically determining the weight ratio based on the scenario, the link switching decision score of the target tire crane can be made more realistic, thereby improving the accuracy of the link switching decision score.

[0058] 205. Determine the coverage area of ​​each base station in the base station group except for the original base station, and select candidate base stations in the base station group based on the coverage area and predicted spatial location.

[0059] Among them, the base station cluster is a 60GHz millimeter wave base station cluster.

[0060] In this step, candidate base station selection involves filtering base stations from a 60GHz millimeter-wave base station cluster that cover the predicted spatial location of the tire crane. The selection is based on the spatial mapping relationship between the base station coverage area and the predicted spatial location, and the selection tool can be the GiNMS platform. The base station cluster consists of 12 60GHz millimeter-wave base stations, deployed on top of 4 lighthouses in the yard (3 base stations per lighthouse). Each base station can use a phase-reconfigurable array antenna, with a horizontal coverage angle of 120°, a vertical coverage angle of 60°, and a coverage radius of 150m, supporting dynamic beamwidth adjustment (30° / 60° / 90°). The GiNMS platform converts the coverage area of ​​each base station into a "polygonal region" in the local coordinate system of the yard and correlates it with real-time occlusion data to correct the coverage boundary. For example, if container stacks obstruct the coverage, the coverage radius on the northeast side of the base station is reduced to 120m. The GiNMS platform uses a "point-polygon collision detection algorithm" to determine whether the predicted location is within the coverage area of ​​the base station, thereby identifying all base stations covering the predicted spatial location from the base station cluster, excluding the original base stations, as candidate base stations.

[0061] 206. Obtain the historical handover success rate for each candidate base station and determine whether the highest historical handover success rate is higher than the preset success rate threshold.

[0062] In this step, the link handover log library of the GiNMS platform is used. This library records information such as handover time, location, scenario (weather, obstruction), and success status for each base station. The historical handover success rate for each candidate base station is calculated by the ratio of the number of successful handovers to the total number of handovers. A preset success rate threshold is used to determine whether the historical handover success rate of the candidate base station is reliable enough to ensure communication continuity and stability when the target tire crane performs link pre-handover, such as 98% or 99%. The optimal solution that meets the remote control requirements can be found through multiple tests, specifically by monitoring the signal interruption rate after base station handover. If the highest historical handover success rate is higher than the preset success rate threshold, step 207 is executed; otherwise, step 208 is executed.

[0063] 207. Select the candidate base station with the highest historical handover success rate as the target base station.

[0064] In this step, the candidate base station with the highest historical handover success rate is directly identified as the target base station to ensure reliability after handover. The central control unit issues a pre-handover preparation command to the target base station, simultaneously pushing information such as the predicted spatial location and operating scenario of the target tire crane. The target base station then enters a beam pre-alignment standby state. Furthermore, candidate base stations with the second-highest historical handover success rate after the target base station can be marked as backup target base stations. If the target base station's signal verification fails subsequently, the system can directly switch to this backup target base station without re-selection.

[0065] As a further implementation method, in order to avoid the risk of handover failure caused by the weak signal penetration capability of millimeter-wave base stations in high-obstruction scenarios and further ensure the reliability of link pre-handover, after selecting the candidate base station with the highest historical handover success rate as the target base station, the method further includes: if the predicted obstruction risk probability is higher than a preset probability threshold, then the hot standby base station is selected as the target base station.

[0066] In this implementation, a preset probability threshold can be set in advance. This threshold is a quantitative standard for determining whether the predicted obstruction risk has reached the critical state of millimeter-wave link failure. Based on the propagation characteristics of 60GHz millimeter-wave signals and measured data from port yard obstruction scenarios, it can be set to 80%. This value range can be fine-tuned according to the port obstruction density, with a typical range of 75%-85%. After determining the candidate base station with the highest historical handover success rate as the target base station, the predicted obstruction risk probability of the target base station is judged. If it is higher than the preset probability threshold, it indicates that after the target tire crane reaches the predicted spatial position, the obstruction of the millimeter-wave signal by surrounding obstacles will cause signal attenuation. Even if the historical handover success rate of the millimeter-wave base station meets the standard, the actual signal strength after handover may still be lower than the preset transmission requirements. The diffraction and penetration capability of the 5.8GHz hot standby base station can resist this level of obstruction. Therefore, the hot standby base station can be used as the target base station.

[0067] 208. Use the hot standby base station as the target base station.

[0068] Among them, the hot standby base station is a broadband wireless base station compatible with the existing 5.8GHz base stations in the port.

[0069] In this step, if any candidate base station fails to qualify as the target base station based on historical handover success rates, the hot standby base station activation process is triggered. This involves using the hot standby base station as the target base station to ensure redundancy during link handover. Specifically, the hot standby base station can consist of three 5.8GHz broadband wireless base stations deployed at the four corners of the yard, covering a 500m radius. These base stations support GPS clock synchronization and fast roaming handover, and are fully compatible with existing 5.8GHz equipment in the port. The central control unit directly triggers the hot standby base station activation command without re-selection (the hot standby base station covers the entire yard). The base station closest to the predicted location is selected by default; if there is only one, selection is unnecessary. The 5.8GHz band has strong diffraction capabilities, maintaining high signal strength even in heavy rain or high-obstruction scenarios, making it a reliable backup for the target tire crane's millimeter-wave link.

[0070] 209. Pre-align the beam direction of the target base station and cover the predicted spatial location.

[0071] This step combines the description of step 104 in the above method, and the same content will not be repeated here.

[0072] 210. Once the target tire crane reaches the predicted spatial location and the signal strength and transmission delay of the target base station are verified to meet the preset transmission requirements, the communication link of the target tire crane is switched from the original base station to the target base station.

[0073] This step combines the description of step 105 in the above method, and the same content will not be repeated here.

[0074] Furthermore, as a further implementation method, to address the risk of communication interruption caused by sudden signal attenuation / blockage of the target base station after handover, thereby ensuring the stability of link handover, after switching the communication link of the target tire crane from the original base station to the target base station, the method further includes: controlling the original base station to remain in standby mode, and monitoring whether the signal strength and transmission delay of the target base station continuously meet the preset transmission requirements within a preset time period; if so, controlling the original base station to shut down the standby mode; if not, switching the communication link of the target tire crane from the target base station back to the original base station or a hot standby base station.

[0075] In this implementation, the standby state is a transitional state between normal operation and complete shutdown of the original base station, which must meet the dual requirements of rapid activation and handover and low resource consumption. The preset monitoring period can be set to 100ms. If the target base station signal is stable after the handover, the index fluctuation within the preset monitoring period is ≤±2dBm. If there is a sudden obstruction / interference, the signal attenuation will appear within 30-50ms within the preset monitoring period, that is, fault identification and handover preparation can be completed within 100ms.

[0076] Furthermore, as a response to the above Figure 1-2 The implementation of the method embodiment shown in this application provides a link pre-switching device based on spatial location. This device is used to reduce link switching latency and avoid signal interruption, meeting the low latency and high reliability requirements of remote control of rubber-tired cranes. The embodiment of this device corresponds to the aforementioned method embodiment. For ease of reading, this embodiment will not repeat the details of the aforementioned method embodiment, but it should be understood that the device in this embodiment can implement all the contents of the aforementioned method embodiment. It is applied to a port rubber-tired crane wireless communication system, specifically as follows... Figure 3 As shown, the device includes: Acquisition unit 301 is used to acquire multi-source operation data of the target tire crane; Prediction unit 302 is used to predict the predicted spatiotemporal operating state of the target tire crane within a specified future time period based on the multi-source operation data obtained by acquisition unit 301 and through a spatiotemporal joint model. The predicted spatiotemporal operating state includes predicted spatial location, predicted attitude parameters and predicted occlusion risk probability. The calculation unit 303 is used to calculate the link switching decision score of the target tire crane based on the predicted spatiotemporal operating status obtained by the prediction unit 302 through a multi-dimensional decision matrix. The decision factors of the multi-dimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical switching success rate and weather attenuation compensation coefficient. Triggering unit 304 is used to trigger a pre-handover operation if the link handover decision score obtained by the calculation unit 303 reaches a preset score threshold. The pre-handover operation includes at least selecting a corresponding target base station in the base station group according to the predicted spatial location and the historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location. The first control unit 305 is used to switch the communication link of the target tire crane from the original base station to the target base station obtained by the triggering unit 304 after the target tire crane reaches the predicted spatial position and verifies that the signal strength and transmission delay of the target base station meet the preset transmission requirements.

[0077] Furthermore, such as Figure 4 As shown, the device further includes: The tagging unit 306 is used to preprocess the multi-source operation data before the prediction unit 302, and add a tag corresponding to the actual spatiotemporal operation status to the preprocessed multi-source operation data within a specified historical time period. The actual spatiotemporal operation status includes the actual spatial location, actual attitude parameters and actual occlusion risk probability. The processing unit 307 is used to perform weighted fusion on the multi-source job data with added tags obtained by the tagging unit 306, and extract the corresponding time series features to obtain time series data; Training unit 308 is used to construct and train a bidirectional LSTM network corresponding to the target tire crane using the time series data obtained by processing unit 307 as samples, so as to obtain the spatiotemporal joint model. The spatiotemporal joint model is used to output the predicted spatiotemporal operating state of the target tire crane in a specified future time period.

[0078] Furthermore, such as Figure 4 As shown, the computing unit 303 includes: The acquisition module 3031 is used to acquire the single-factor scoring rules corresponding to each of the decision factors, and calculate the single-factor score of each of the decision factors based on the single-factor scoring rules. The calculation module 3032 is used to perform weighted summation of the single factor scores of each decision factor obtained by the acquisition module 3031 based on a preset weight allocation rule to obtain the link switching decision score. The weight allocation rule is dynamically changed according to the different port scenarios corresponding to the target rubber-tired crane. The different port scenarios include high-speed transfer scenarios, rainy day operation scenarios, core control operation scenarios, and high obstruction scenarios.

[0079] Furthermore, such as Figure 4 As shown, the calculation module 3032 specifically includes, Obtain the initial weight percentage of different decision factors in the weight allocation rule; When in the high-speed transition scenario, the initial weight ratio of the signal attenuation slope is increased according to a preset gradient, and the initial weight ratio of the historical handover success rate is decreased according to a preset gradient. When operating in the rainy weather scenario, the initial weight of the weather attenuation compensation coefficient is increased according to a preset gradient, and the initial weight of the shading risk index is decreased according to a preset gradient. When in the core control operation scenario, the weight ratio of the service priority is increased according to the preset gradient, and the initial weight ratios of the signal attenuation slope, the obstruction risk index, the historical handover success rate and the weather attenuation compensation coefficient are proportionally distributed. When in the high obstruction scenario, the initial weight ratio of the obstruction risk index is increased according to the preset gradient, the initial weight ratio of the historical handover success rate is increased according to the preset gradient, and the initial weight ratios of the signal attenuation slope, the service priority, and the weather attenuation compensation coefficient are proportionally allocated.

[0080] Furthermore, such as Figure 4 As shown, the triggering unit 304 includes: The filtering module 3041 is used to determine the coverage area of ​​each base station in the base station group except for the original base station, and to filter candidate base stations in the base station group based on the coverage area and the predicted spatial location. The base station group is a 60GHz millimeter wave base station cluster. The judgment module 3042 is used to obtain the historical handover success rate corresponding to each candidate base station obtained by the filtering module 3041, and to determine whether the highest historical handover success rate is higher than a preset success rate threshold. The first determining module 3043 is used to determine that if the judging module 3042 determines that the highest historical handover success rate is higher than a preset success rate threshold, the candidate base station with the highest historical handover success rate is taken as the target base station. The second determining module 3044 is used to determine the target base station if the judgment module 3043 determines that the highest historical handover success rate is not higher than a preset success rate threshold. The hot standby base station is a broadband wireless base station compatible with the existing 5.8GHz base stations in the port. The control module 3045 is used to control the beam direction of the target base station obtained by the first determining module 3043 or the second determining module 3044 to be pre-aligned and cover the predicted spatial location.

[0081] Furthermore, such as Figure 4 As shown, the device further includes: The third determining module 3046 is used to designate the hot standby base station as the target base station if the predicted occlusion risk probability is higher than a preset probability threshold after the first determining module 343.

[0082] Furthermore, such as Figure 4 As shown, the device further includes: The monitoring unit 309 is used to control the original base station to remain in standby state after the first control unit 305, and to monitor whether the signal strength and transmission delay of the target base station continuously meet the preset transmission requirements within a preset time period. The second control unit 310 is configured to control the original base station to turn off the standby state if the monitoring unit 309 monitors the signal strength and transmission delay of the target base station within a preset time period and continuously meets the preset transmission requirements. The third control unit 311 is configured to switch the communication link of the target tire crane from the target base station back to the original base station or the hot standby base station if the monitoring unit 309 monitors the signal strength and transmission delay of the target base station within a preset time period and finds that they do not continuously meet the preset transmission requirements.

[0083] Furthermore, embodiments of this application also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The spatial location-based link pre-switching method described in the article.

[0084] Furthermore, embodiments of this application also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The spatial location-based link pre-switching method described in the article.

[0085] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0086] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0087] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0088] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this application is not directed to any particular programming language. It should be understood that the content of this application described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of this application.

[0089] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0090] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0092] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0094] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0095] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0096] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0097] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0098] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0099] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A link pre-switching method based on spatial location, applied to a port rubber-tired gantry crane wireless communication system, characterized in that, The method includes: Acquire multi-source operational data of the target tire crane; Based on the multi-source operation data, the spatiotemporal operating status of the target tire crane within a specified future time period is predicted by a spatiotemporal joint model. The predicted spatiotemporal operating status includes the predicted spatial location, predicted attitude parameters, and predicted occlusion risk probability. Based on the predicted spatiotemporal operating status, the link switching decision score of the target tire crane is calculated through a multidimensional decision matrix. The decision factors of the multidimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical switching success rate and weather attenuation compensation coefficient. If the link handover decision score reaches a preset score threshold, a pre-handover operation is triggered. The pre-handover operation includes at least selecting a corresponding target base station in the base station group based on the predicted spatial location and the historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location. Once the target tire crane reaches the predicted spatial location and verifies that the signal strength and transmission delay of the target base station meet the preset transmission requirements, the communication link of the target tire crane is switched from the original base station to the target base station.

2. The method according to claim 1, characterized in that, Before predicting the spatiotemporal operating status of the target tire crane in a future period using a spatiotemporal joint model based on the multi-source operation data, the method further includes: The multi-source operation data is preprocessed, and the preprocessed multi-source operation data within a specified historical time period is labeled with tags corresponding to the actual spatiotemporal operation status, which includes the actual spatial location, actual attitude parameters, and actual occlusion risk probability. The multi-source job data with added tags are weighted and fused, and the corresponding time series features are extracted to obtain time series data; Using the time-series data as samples, a bidirectional LSTM network corresponding to the target tire crane is constructed and trained to obtain the spatiotemporal joint model. The spatiotemporal joint model is used to output the predicted spatiotemporal operating state of the target tire crane within a specified future time period.

3. The method according to claim 1, characterized in that, Based on the predicted spatiotemporal operating state, the link switching decision score of the target tire crane is calculated using a multi-dimensional decision matrix, including: Obtain the single-factor scoring rules corresponding to each decision factor, and calculate the single-factor score of each decision factor based on the single-factor scoring rules. Based on the preset weight allocation rules, the single-factor scores of each decision factor are weighted and summed to obtain the link switching decision score. The weight allocation rules are dynamically changed according to the different port scenarios corresponding to the target rubber-tired gantry crane. The different port scenarios include high-speed transfer scenarios, rainy day operation scenarios, core control operation scenarios, and high obstruction scenarios.

4. The method according to claim 3, characterized in that, Based on a preset weight allocation rule, the single-factor scores of each decision factor are weighted and summed to obtain the link switching decision score, including: Obtain the initial weight percentage of different decision factors in the weight allocation rule; When in the high-speed transition scenario, the initial weight ratio of the signal attenuation slope is increased according to a preset gradient, and the initial weight ratio of the historical handover success rate is decreased according to a preset gradient. When operating in the rainy weather scenario, the initial weight of the weather attenuation compensation coefficient is increased according to a preset gradient, and the initial weight of the shading risk index is decreased according to a preset gradient. When in the core control operation scenario, the weight ratio of the service priority is increased according to the preset gradient, and the initial weight ratios of the signal attenuation slope, the obstruction risk index, the historical handover success rate and the weather attenuation compensation coefficient are proportionally distributed. When in the high obstruction scenario, the initial weight ratio of the obstruction risk index is increased according to the preset gradient, the initial weight ratio of the historical handover success rate is increased according to the preset gradient, and the initial weight ratios of the signal attenuation slope, the service priority, and the weather attenuation compensation coefficient are proportionally allocated.

5. The method according to claim 1, characterized in that, Triggering pre-switching operations includes: The coverage area of ​​each base station in the base station cluster, excluding the original base station, is determined, and candidate base stations are selected in the base station cluster based on the coverage area and the predicted spatial location. The base station cluster is a 60GHz millimeter wave base station cluster. Obtain the historical handover success rate corresponding to each candidate base station, and determine whether the highest historical handover success rate is higher than a preset success rate threshold; If so, the candidate base station with the highest historical handover success rate will be selected as the target base station; If not, then the hot standby base station will be used as the target base station, and the hot standby base station is a broadband wireless base station compatible with the existing 5.8GHz base stations in the port. The beam direction of the target base station is pre-aligned and covers the predicted spatial location.

6. The method according to claim 5, characterized in that, After selecting the candidate base station with the highest historical handover success rate as the target base station, the method further includes: If the predicted occlusion risk probability is higher than a preset probability threshold, then the hot standby base station will be used as the target base station.

7. The method according to any one of claims 1-6, characterized in that, After switching the communication link of the target tire crane from the original base station to the target base station, the method further includes: The original base station is kept in standby mode, and the signal strength and transmission delay of the target base station are monitored within a preset time period to ensure that they continuously meet the preset transmission requirements. If so, then control the original base station to turn off the standby state; If not, the communication link of the target tire crane will be switched from the target base station back to the original base station or the hot standby base station.

8. A link pre-switching device based on spatial location, characterized in that, The device, used in a port rubber-tired gantry crane wireless communication system, includes: The acquisition unit is used to acquire multi-source operation data of the target tire crane; The prediction unit is used to predict the predicted spatiotemporal operating state of the target tire crane within a specified future time period based on the multi-source operation data obtained by the acquisition unit and through a spatiotemporal joint model. The predicted spatiotemporal operating state includes the predicted spatial location, predicted attitude parameters, and predicted occlusion risk probability. The calculation unit is used to calculate the link switching decision score of the target tire crane based on the predicted spatiotemporal operating status obtained by the prediction unit through a multi-dimensional decision matrix. The decision factors of the multi-dimensional decision matrix include signal attenuation slope, obstruction risk index, service priority, historical switching success rate and weather attenuation compensation coefficient. A triggering unit is used to trigger a pre-handover operation if the link handover decision score obtained by the calculation unit reaches a preset score threshold. The pre-handover operation includes at least selecting a corresponding target base station in the base station group based on the predicted spatial location and the historical handover success rate, and controlling the beam direction of the target base station to be pre-aligned and cover the predicted spatial location. The first control unit is used to switch the communication link of the target tire crane from the original base station to the target base station obtained by the triggering unit after the target tire crane reaches the predicted spatial position and verifies that the signal strength and transmission delay of the target base station meet the preset transmission requirements.

9. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the spatial location-based link pre-switching method as described in any one of claims 1 to 7.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the spatial location-based link pre-switching method as described in any one of claims 1 to 7.

Images