A method for solving the multi-source pose fusion of a mast crane operation end based on Beidou

Abstract

The present application relates to the technical field of engineering machinery pose detection, and discloses a mast crane operation end multi-source pose fusion solving method based on Beidou, comprising: synchronously collecting Beidou positioning, IMU inertial navigation, winch encoder and jib inclination data, outputting standardized time series data set and sensor calibration initial value after preprocessing; respectively performing Beidou carrier wave phase difference solution and strapdown inertial navigation recursion, outputting double-source pose and corresponding confidence; using adaptive federated Kalman filtering to complete multi-source initial fusion, outputting initial fusion pose and fusion residual error; introducing an operation end kinematics constraint model to correct errors, outputting final pose solution result and precision evaluation parameters; according to the solution result and residual error, iteratively optimizing calibration parameters and filtering gain to form a closed loop self-optimization. The present application realizes high-precision real-time solution of the mast crane operation end pose, effectively suppresses Beidou shielding and inertial navigation drift error, and improves the solution robustness and long-term stability.

Description

Technical Field

[0001] This invention relates to the field of engineering machinery pose detection technology, and in particular to a multi-source pose fusion calculation method for the mast crane working end based on Beidou. Background Technology

[0002] As a core lifting equipment in ports, terminals, and construction sites, the real-time position and orientation of the spreader at the working end of a mast crane are key parameters for achieving precise lifting, collision avoidance, and automated operation. Traditional mast crane position and orientation detection methods rely heavily on winch encoders and limit switches, which can only obtain rough rope length and boom angle data, failing to provide the three-dimensional absolute position and orientation of the working end, and suffer from low accuracy and poor real-time performance.

[0003] Existing solutions incorporate BeiDou positioning technology to obtain the boom position, but a single BeiDou system is prone to loss of lock or accuracy degradation in scenarios with obstruction or multiple paths, failing to provide consistently reliable pose. Combined solutions with inertial navigation often employ simple Kalman filtering fusion without adaptively adjusting weights based on signal quality, leading to rapid accumulation of inertial navigation drift errors when BeiDou loses lock. Furthermore, most existing solutions only calculate the boom tip pose, failing to incorporate winch rope length and boom tilt angle to construct kinematic constraints for the work end, thus failing to directly obtain the precise pose of the spreader end and lacking a self-optimization mechanism for calibration parameters. Over long-term operation, installation offsets and boom deformation cause a continuous decline in calculation accuracy. Therefore, a work end pose calculation method is needed that can achieve adaptive fusion of BeiDou and inertial navigation, incorporate kinematic constraint correction, and possess closed-loop self-optimization capabilities. Summary of the Invention

[0004] To address the aforementioned technical shortcomings, the purpose of this invention is to propose a multi-source pose fusion calculation method for mast crane operation terminals based on BeiDou. This method aims to solve the technical problems in existing technologies, such as poor reliability of single BeiDou, lack of inertial navigation drift suppression, lack of kinematic constraint correction for operation terminals, and lack of closed-loop self-optimization mechanism, which lead to low pose calculation accuracy, poor scene adaptability, and insufficient long-term stability.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a multi-source pose fusion calculation method for mast crane operation end based on Beidou.

[0006] The aforementioned multi-source pose fusion calculation method for a mast crane operating end based on BeiDou includes: Step S10: Synchronously collect BeiDou positioning data, IMU inertial navigation data, winch encoder data and boom tilt angle sensing data of the mast boom. After outlier removal, spatiotemporal alignment and coordinate system preprocessing, output standardized multi-source time series dataset, unified coordinate system transformation parameters and sensor calibration initial values. Step S20: Based on the standardized multi-source time series dataset obtained in step S10, perform BeiDou carrier phase difference decomposition and strapdown inertial navigation recursive calculation, and output BeiDou high-precision pose, inertial navigation recursive pose and corresponding confidence index. Step S30: Based on the initial sensor calibration values ​​from step S10 and the BeiDou high-precision pose, inertial navigation recursive pose, and confidence index output from step S20, an adaptive federated Kalman filter is used to perform initial fusion of multi-source poses, and the initial fusion pose result and fusion residual sequence are output. Step S40: Based on the initial fusion pose result output in step S30 and the winch encoder data and boom tilt angle sensing data in step S10, the kinematic constraint model of the working end is introduced to perform error correction, and the final pose calculation result and pose accuracy evaluation parameters of the working end are output. Step S50: Based on the final pose calculation result output in step S40 and the fusion residual sequence output in step S30, perform iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, output the pose calculation steady-state flag and feed back to update the calculation baseline.

[0007] Preferably, step S10, which involves simultaneously acquiring BeiDou positioning data, IMU inertial navigation data, winch encoder data, and boom tilt sensor data from the mast boom, and performing outlier removal, spatiotemporal alignment, and coordinate system preprocessing to obtain a standardized multi-source time-series dataset, unified coordinate system transformation parameters, and initial sensor calibration values, specifically includes: Step S101: Collect raw observation data and ephemeris data output by the Beidou dual-mode receiver installed at the top of the boom at a sampling rate of 1Hz. At the same time, collect triaxial acceleration and triaxial angular velocity data output by the IMU installed at the same position at a sampling rate of 200Hz. Collect rope length data output by the winch shaft encoder and pitch angle data output by the boom tilt sensor at a sampling rate of 50Hz. Step S102: Perform multipath suppression and outlier removal on the original BeiDou observation data, perform fourth-order Butterworth low-pass filtering on the IMU data with a cutoff frequency of 20Hz, perform sliding mean denoising on the encoder and tilt data, and interpolate and align all sensor data with the BeiDou second pulse as the time reference to achieve millisecond-level spatiotemporal synchronization and generate a standardized multi-source time series dataset with equal intervals. Step S103: Establish the transformation relationship between the BeiDou coordinate system, the inertial navigation coordinate system and the mast crane body coordinate system, calibrate the installation deviation of each sensor and the geometric parameters of the boom, and output the unified coordinate system transformation parameters and the initial calibration values ​​of the sensors.

[0008] Preferably, step S20, based on the standardized multi-source time-series dataset obtained in step S10, involves performing BeiDou carrier phase difference decomposition and strapdown inertial navigation recursive calculation to output high-precision BeiDou pose, inertial navigation recursive pose, and corresponding confidence index. This step specifically includes: Step S201: Based on the original BeiDou observation data and the differential data of the base station, the carrier phase differential algorithm is used for calculation. Combined with cycle slip detection and ambiguity fixing, the BeiDou high-precision positioning result and attitude angle at the top of the boom are obtained. At the same time, the BeiDou positioning reliability is calculated based on the number of visible stars and the PDOP value. Step S202: Using the initial sensor calibration value from step S10 as the initial alignment parameter, perform strapdown inertial navigation recursive calculation on the filtered IMU data, output the position, velocity and attitude angle in the inertial navigation system, and calculate the inertial navigation confidence based on the recursive duration and angular velocity noise level. Step S203: Convert the BeiDou pose and inertial navigation pose to the body coordinate system, and output the BeiDou high-precision pose, the inertial navigation recursive pose and their respective confidence indexes. The confidence value ranges from 0 to 1.

[0009] Preferably, in step S201, the BeiDou positioning confidence level is graded according to the number of visible stars and the PDOP value. When the number of visible stars is ≥8 and the PDOP is ≤2, the confidence level is 0.95; when the number of visible stars is 5~7 and the PDOP is ≤4, the confidence level is 0.7; when the number of visible stars is <5, the confidence level is 0.3 or below; and when the confidence level is >0.

[0010] Preferably, step S30, based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index, and sensor calibration initial values ​​output in step S20, employs adaptive federated Kalman filtering to perform initial fusion of multi-source poses, obtaining the initial fused pose result, fusion residual sequence, and kinematic constraint parameter set, specifically includes: Step S301: Using the initial sensor calibration value as the coordinate system reference, and using BeiDou and inertial navigation as two independent sub-filters, each with its own confidence level as the observation noise weight coefficient, adaptively adjust the observation noise covariance matrix of the sub-filters. Step S302: The federated Kalman filter architecture is adopted. The two sub-filters are updated in parallel. The main filter performs information fusion according to confidence weight to obtain the globally optimal initial fused pose result of the boom top. Step S303: Record the new information sequence of the filter update at each time step as the fusion residual. Simultaneously, after time stamp matching, the basic kinematic constraint parameters are encapsulated into a kinematic constraint parameter set. Output the initial fusion pose result, the fusion residual sequence, and the kinematic constraint parameter set.

[0011] Preferably, step S40, based on the initial fused pose result and kinematic constraint parameter set output in step S30, introduces the execution error correction of the kinematic constraint model at the working end to obtain the final pose calculation result and pose accuracy evaluation parameters at the working end, specifically includes: Step S401: Extract boom tilt angle data and winch rope length data from the kinematic constraint parameter set, combine them with boom geometric parameters, construct a kinematic model of the mast crane working end, and calculate the theoretical pose of the working end spreader from the initial fused pose of the boom top. Step S402: Introduce rope length constraints and swing angle constraints to correct kinematic errors in the initial fusion pose, compensate for pose deviations caused by boom deformation and rope swing, and obtain the three-dimensional position and attitude angle of the lifting device at the working end. Step S403: Calculate pose accuracy evaluation parameters based on the fusion residual and constraint deviation, including horizontal position error estimation and elevation error estimation, and output the final pose calculation result and pose accuracy evaluation parameters of the working end.

[0012] Preferably, in step S50, the steps of performing iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, outputting a steady-state flag of pose calculation and feeding back to update the calculation baseline based on the final pose calculation result output in step S40 and the fusion residual sequence output in step S30, specifically include: Step S501: When the BeiDou confidence level is higher than the set threshold and the fused residual sequence remains stable, the final pose calculation result is used as the comparison benchmark. The sensor installation deviation and boom geometric parameters are iteratively corrected by combining the high-confidence BeiDou observation values. The optimized calibration parameters are output and fed back to step S10 to update the benchmark. Step S502: Calculate the observation noise bias based on the statistical characteristics of the fused residual sequence, and estimate the process noise level by combining the temporal fluctuation of the final pose solution result. Then, jointly and adaptively adjust the process noise covariance and observation noise weight of the federated Kalman filter to optimize the filter gain of subsequent frames. Step S503: When the fusion residuals of 30 consecutive frames are all less than the set residual threshold, and the pose fluctuation of the final pose calculation result is less than the limit, the calculation is determined to be converged and the pose calculation steady-state flag is set, the steady-state flag is output and the initial baseline of the calculation in step S20 is updated.

[0013] This invention also provides a multi-source pose fusion calculation system for a mast crane operating end based on BeiDou, comprising: Multi-source sensor acquisition and preprocessing module: used to simultaneously acquire BeiDou positioning data, IMU inertial navigation data, winch encoder data and boom tilt angle sensor data of mast crane boom. After outlier removal, spatiotemporal alignment and coordinate system preprocessing, it outputs standardized multi-source time series dataset, unified coordinate system transformation parameters and sensor calibration initial values. Dual-source pose independent solution module: Based on the standardized multi-source time series dataset and sensor calibration initial values ​​obtained by the multi-source sensor acquisition preprocessing module, it performs BeiDou carrier phase difference solution and strapdown inertial navigation recursive solution, and outputs BeiDou high-precision pose, inertial navigation recursive pose, corresponding confidence index and kinematic constraint basic parameters. Adaptive filtering initial fusion module: Based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index and sensor calibration initial value output by the dual-source pose independent solution module, the module uses adaptive federated Kalman filtering to perform multi-source pose initial fusion, and outputs the initial fusion pose result, fusion residual sequence and kinematic constraint parameter set. Kinematics constraint correction module: Based on the initial fusion pose result and kinematic constraint parameter set output by the adaptive filtering initial fusion module, it introduces the kinematic constraint model of the working end to perform error correction, and outputs the final pose solution result and pose accuracy evaluation parameters of the working end; Iterative optimization closed-loop feedback module: Based on the final pose calculation result output by the kinematic constraint correction module and the fusion residual sequence output by the adaptive filtering initial fusion module, it performs iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, outputs pose calculation steady-state flag and feeds back to update the calculation baseline.

[0014] The present invention also provides a BeiDou-based mast crane operation end multi-source pose fusion calculation device, the BeiDou-based mast crane operation end multi-source pose fusion calculation device comprising: a memory, a processor, and a BeiDou-based mast crane operation end multi-source pose fusion calculation program stored in the memory and executable on the processor, wherein the BeiDou-based mast crane operation end multi-source pose fusion calculation program implements the above method when executed by the processor.

[0015] The present invention also provides a computer program product, which includes a BeiDou-based multi-source pose fusion calculation program for mast crane operation end, wherein the BeiDou-based multi-source pose fusion calculation program for mast crane operation end implements the above-mentioned method when executed by a processor.

[0016] The beneficial effects of this invention are as follows: By synchronously collecting multi-source data from BeiDou, IMU, encoder, and tilt angle, and completing spatiotemporal alignment and coordinate system unification, a multi-dimensional pose observation base was constructed. This broke through the observation limitations of a single sensor and provided multi-source redundant data support for subsequent layered fusion, improving the reliability of the solution from the data source level.

[0017] A confidence-adaptive federated Kalman filter architecture is adopted, which dynamically adjusts the fusion weights according to the quality of the BeiDou signal and the degree of inertial navigation drift. When the BeiDou signal is good, BeiDou is used as the main source to correct the inertial navigation drift. When BeiDou is blocked, inertial navigation is used as the main source to maintain the pose output. This approach balances high accuracy and high continuity, effectively improving the robustness of pose calculation in complex scenarios.

[0018] By introducing a kinematic constraint model of the winch rope length and boom tilt angle, the boom top pose is further modified into the working end spreader pose, directly outputting the end pose required for lifting operations. At the same time, it compensates for systematic errors caused by boom deformation and rope swing, making the solution results more in line with actual operation requirements.

[0019] By integrating residuals and iteratively optimizing calibration parameters and filter gains, a complete closed loop of acquisition, calculation, fusion, correction, and optimization is formed. This can automatically calibrate parameter drift caused by installation deviations and long-term deformation, enabling the system to maintain accuracy over long-term operation without the need for frequent manual calibration. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the first embodiment of a multi-source pose fusion calculation method for a mast crane operating end based on BeiDou according to the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0023] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of a multi-source pose fusion calculation method for a mast crane operation end based on BeiDou according to the present invention. The first embodiment of the multi-source pose fusion calculation method for a mast crane operation end based on BeiDou according to the present invention is presented.

[0024] In the first embodiment, the multi-source pose fusion calculation method for a mast crane operating end based on BeiDou includes: Step S10: Synchronously collect BeiDou positioning data, IMU inertial navigation data, winch encoder data and boom tilt angle sensing data of the mast boom. After outlier removal, spatiotemporal alignment and coordinate system preprocessing, output standardized multi-source time series dataset, unified coordinate system transformation parameters and sensor calibration initial values.

[0025] It should be noted that in this step, the BeiDou receiver is installed at the top of the mast boom to obtain the absolute position reference; the IMU is co-located with the BeiDou receiver to provide high-frequency motion attitude observation; the winch encoder is installed at the end of the winch drum shaft to measure the wire rope release length; and the tilt sensor is installed at the boom root to collect the boom pitch angle in real time. Outlier removal uses the 3σ criterion to remove BeiDou observation jumps and encoder pulse interference; low-pass filtering suppresses the mechanical vibration noise of the IMU; spatiotemporal alignment uses the BeiDou second pulse as a unified time reference, and linear interpolation is performed on sensor data with different sampling rates to ensure that all observation data strictly correspond on the time axis. Coordinate system one completes the calibration of the transformation matrix between the BeiDou geodetic coordinate system, the inertial navigation carrier coordinate system, and the mast crane body coordinate system, and provides the initial calibration values ​​for geometric parameters such as sensor installation deviation and boom length. The output standardized time series dataset is used for independent calculation in step S20, while the coordinate system transformation parameters and initial calibration values ​​provide a reference for subsequent fusion and kinematic calculations.

[0026] Understandably, through this preprocessing step, heterogeneous, frequency-dependent, and coordinate-system-dependent multi-source sensor data are integrated into a standardized dataset that is spatiotemporally unified and has a consistent benchmark. This eliminates the fusion errors caused by time misalignment and coordinate system deviation, laying a reliable data foundation for subsequent high-precision pose calculation.

[0027] It should be understood that traditional multi-sensor fusion schemes often ignore the impact of spatiotemporal alignment accuracy and calibration errors, directly performing data fusion, which leads to system errors being carried into the final result. This step, through rigorous spatiotemporal alignment and initial calibration, controls system errors from the source, ensuring the consistency of the benchmark for subsequent fusion solutions.

[0028] For example, a Beidou receiver and IMU are installed at the top of the boom of a port mast, an encoder is installed at the winch end, and an angle sensor is installed at the base of the boom. The sampling rates of each sensor are 1Hz, 200Hz, and 50Hz, respectively. After processing in this step, all data are uniformly aligned to the 10Hz output sequence, the coordinate system is changed to the local horizontal coordinate system, and the IMU installation angle and the effective length of the boom are calibrated, providing a unified time and space reference for subsequent calculations.

[0029] Step S20: Based on the standardized multi-source time series dataset obtained in step S10, perform BeiDou carrier phase difference decomposition and strapdown inertial navigation recursive calculation, and output BeiDou high-precision pose, inertial navigation recursive pose and corresponding confidence index.

[0030] It should be noted that this step adopts a dual-source independent solution architecture. The BeiDou end achieves centimeter-level positioning through carrier phase differential technology, and combines cycle slip detection and integer ambiguity fixing algorithms to ensure positioning accuracy. The confidence level is comprehensively evaluated based on the number of visible satellites, spatial geometric distribution, and PDOP value, reflecting the reliability of the current BeiDou solution. The inertial navigation end is based on the strapdown inertial navigation principle, using acceleration and angular velocity for integral recursion to output high-frequency position, velocity, and attitude. The inertial navigation confidence level decreases as the recursion time increases, and is also corrected by incorporating angular velocity noise level to reflect the degree of inertial navigation drift. The two sets of solution results are uniformly converted to the body coordinate system and output, providing two independent pose observations for subsequent fusion. The output BeiDou and inertial navigation pose and confidence levels are input into step S30 for fusion processing.

[0031] Understandably, this step leverages the advantages of BeiDou's long-term high precision and inertial navigation's short-term high dynamics to output two independent poses and confidence levels. This preserves the observation characteristics of each device and provides a weighting basis for subsequent adaptive fusion, thus avoiding the failure of the overall solution due to the failure of a single sensor.

[0032] It should be understood that traditional integrated navigation schemes often directly use BeiDou as the input for Kalman filtering of observations without quantitatively evaluating the quality of BeiDou itself. Signal degradation can then contaminate the filtering results. This step introduces a confidence index to quantify the reliability of BeiDou and inertial navigation at each moment, enabling dynamic weight allocation for subsequent fusion and improving the system's anti-interference capability.

[0033] For example, in an open area, there are 10 visible BeiDou satellites, a PDOP value of 1.5, and the ambiguity is successfully fixed, with a positioning accuracy at the centimeter level and a confidence level of 0.95. At this time, the inertial navigation system (INS) has just completed alignment, with very little drift and a confidence level of 0.9. When the boom moves to the area below the gantry crane and is obstructed, the number of visible satellites drops to 4, the PDOP rises to 5, the ambiguity cannot be fixed, and the BeiDou positioning accuracy drops to the meter level, with a confidence level of 0.25. At this time, the short-term recursive inertial navigation system still maintains high accuracy with a confidence level of 0.85, providing reliable high-frequency pose output for subsequent fusion.

[0034] Step S30: Based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index and sensor calibration initial values ​​output in step S20, perform multi-source pose initial fusion using adaptive federated Kalman filtering, and output the initial fusion pose result, fusion residual sequence and kinematic constraint parameter set.

[0035] It should be noted that the federated Kalman filter adopts a "parallel sub-filter + main filter fusion" architecture. BeiDou and inertial navigation are updated independently as sub-filters, avoiding the propagation of single-filter faults. The adaptive mechanism is reflected in the dynamic adjustment of the observation noise covariance with confidence level: high confidence level results in low observation noise and corresponding high weights; low confidence level results in high observation noise and corresponding low weights. The main filter performs weighted fusion of the results of the two sub-filters according to the information allocation principle to obtain the globally optimal boom tip pose. The fusion residual, i.e., the filtering innovation, reflects the degree of deviation between the observed and predicted values ​​and is used for subsequent closed-loop optimization. The output initial fused pose is input into step S40 for kinematic correction, while the fusion residual is input into step S50 for parameter optimization.

[0036] Understandably, this step achieves complementary advantages between BeiDou and inertial navigation through adaptive federated filtering: BeiDou corrects the cumulative drift of inertial navigation, inertial navigation compensates for the sampling gap and brief loss of lock of BeiDou, and confidence weighting ensures that the fusion weights are always tilted towards the more reliable sensor, so that the output pose is superior to that of a single sensor in terms of both accuracy and continuity.

[0037] It should be understood that traditional integrated navigation systems often employ standard Kalman filtering, which results in fixed observation noise and cannot adapt to scenarios with dynamically changing BeiDou signals. This can easily lead to filter divergence when the signal is poor. The adaptive weighting mechanism in this step can adjust the fusion strategy in real time according to signal quality, significantly improving filtering stability under complex operating conditions.

[0038] For example, in open conditions, the confidence level of BeiDou is 0.95 and that of inertial navigation is 0.9. During fusion, BeiDou has a higher weight, and the output pose is based on the centimeter-level accuracy of BeiDou, while the output frequency is increased by utilizing inertial navigation. When BeiDou briefly loses lock, the confidence level drops to 0.2, and the fusion weight automatically tilts towards inertial navigation. The pose output is maintained by recursion from inertial navigation, avoiding positioning jumps. After BeiDou recovers, the weight automatically returns to its original position to correct the inertial navigation drift.

[0039] Based on the initial fusion pose result and kinematic constraint parameter set output in step S30, the execution error correction of the kinematic constraint model at the working end is introduced, and the final pose solution result and pose accuracy evaluation parameters at the working end are output.

[0040] It should be noted that the kinematic model of the working end uses the boom tip pose as a reference, and combines the boom tilt angle, winch rope length, and spreader geometric parameters to calculate the three-dimensional position and attitude of the spreader center point through spatial geometric transformation. Rope length constraints use encoder-measured rope length to correct the spreader elevation, and swing angle constraints use the physical limits of rope swing to limit the horizontal deviation range, thereby compensating for the pose errors caused by boom elastic deformation and rope swing. Accuracy evaluation parameters are based on a comprehensive estimate of fused residuals and constraint deviations, providing the error range in the horizontal and vertical directions, providing a reliable reference for subsequent lifting control. The output final pose of the working end can be directly used for lifting positioning, collision avoidance, and automatic operation control.

[0041] Understandably, this step further calculates the pose of the boom tip to the lifting end, so that the output directly corresponds to the execution end of the lifting operation. At the same time, additional observation constraints such as rope length and tilt angle are introduced through kinematic constraints, which further compresses the error range of the pose calculation and improves the positioning accuracy of the end.

[0042] It should be understood that most existing solutions only output the pose of the boom tip, which cannot directly reflect the actual position of the lifting device, requiring manual conversion in field applications; moreover, they do not consider the effects of boom deformation and rope sway, resulting in actual end-effector error being much greater than boom tip error. The kinematic constraint correction in this step directly outputs the working end pose and reduces system error through multiple constraints, making it more in line with actual engineering needs.

[0043] For example, the initial fusion position of the boom top has a horizontal accuracy of 2cm and an elevation accuracy of 3cm. Combining the 12-meter winch rope length and boom tilt angle data, the spreader's position is calculated through a kinematic model. After introducing rope length constraints, the spreader's elevation error is compressed to within 2cm. After introducing swing angle constraints, the horizontal swing deviation is effectively suppressed. The final output of the spreader's three-dimensional position can meet the accuracy requirements for precise lifting and alignment.

[0044] Step S50: Based on the final pose calculation result output in step S40 and the fusion residual sequence output in step S30, perform iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, output the pose calculation steady-state flag and feed back to update the calculation baseline.

[0045] It should be noted that during the high-confidence period of BeiDou, parameter iterative optimization uses high-precision BeiDou pose as the true reference. The least squares method is used to iteratively correct calibration parameters such as sensor installation deviation and boom length, offsetting the effects of installation errors and long-term deformation. Filter gain adjustment is based on the statistical characteristics of the fusion residual. If the residual remains consistently large, the process noise is increased or the observation weights are adjusted accordingly to maintain optimal filtering performance. The steady-state flag indicates whether the current solution has entered a stable convergence state and can serve as a ready signal for the downstream control system. The optimized calibration parameters are fed back to step S10 to update the baseline, and the filter parameters are fed back to steps S20 and S30 to adjust the solution and fusion strategy, forming a complete self-optimizing closed loop.

[0046] Understandably, this step enables the system to achieve self-calibration and self-optimization capabilities, automatically compensating for parameter drift without the need for regular manual calibration, thus maintaining stable pose calculation accuracy during long-term operation; at the same time, the steady-state flag provides a reliable status indication for subsequent applications, avoiding unreliable pose output when the system has not converged.

[0047] It should be understood that traditional pose calculation systems have fixed calibration parameters. As equipment ages, booms deform, and installations loosen, the calculation error gradually increases, requiring regular manual calibration and maintenance. The closed-loop iterative optimization mechanism in this step enables the system to have self-healing capabilities, significantly reducing maintenance costs and improving long-term operational reliability.

[0048] For example, after three months of system operation, the boom undergoes slight deformation due to load, causing a deviation in the initially calibrated boom length and resulting in a gradual increase in the spreader elevation error. When BeiDou is in an open environment with high confidence, the system automatically uses BeiDou elevation as a reference to iteratively optimize the boom length parameters. After several iterations, the elevation error is restored to the initial accuracy level, eliminating the need for manual measurement and calibration on the boom.

[0049] Example 2: Furthermore, the present invention provides a multi-source pose fusion calculation system for a mast crane operating end based on BeiDou, employing a multi-source pose fusion calculation method for a mast crane operating end based on BeiDou from the above embodiments. This system can solve the technical problems in the prior art, such as poor reliability of a single BeiDou system, lack of inertial navigation drift suppression, lack of kinematic constraint correction for the operating end, and lack of closed-loop self-optimization mechanism, resulting in low pose calculation accuracy, poor scene adaptability, and insufficient long-term stability. The beneficial effects of the multi-source pose fusion calculation system for a mast crane operating end based on BeiDou provided by the present invention are the same as those of the multi-source pose fusion calculation method for a mast crane operating end based on BeiDou provided in the above embodiments. Other technical features of the multi-source pose fusion calculation system for a mast crane operating end based on BeiDou are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0050] Example 3: This invention provides a BeiDou-based multi-source pose fusion calculation device for mast crane operation. The device includes: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which are executed to enable the at least one processor to perform the BeiDou-based multi-source pose fusion calculation method for mast crane operation described in Example 1. The BeiDou-based multi-source pose fusion calculation device for mast crane operation in this invention can include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This BeiDou-based multi-source pose fusion calculation device for mast crane operation is merely an example and should not limit the functionality or scope of the invention. A BeiDou-based mast crane multi-source pose fusion calculation device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes based on programs stored in read-only memory or programs loaded from storage devices into random access memory. The random access memory also stores various programs and data required for the operation of the BeiDou-based mast crane multi-source pose fusion calculation device. The processing unit, read-only memory, and random access memory are interconnected via a bus. An I / O interface is also connected to the bus. Typically, the following systems can be connected to the I / O interface: input devices including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices including, for example, magnetic tapes, hard disks, etc.; and communication devices. The communication device allows the BeiDou-based mast crane multi-source pose fusion calculation device to exchange data with other devices wirelessly or via wired communication. While a multi-source pose fusion calculation device based on BeiDou for mast crane operations has been described, it should be understood that implementation of all the described systems is not required. More or fewer systems may be implemented alternatively.

[0051] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described multi-source pose fusion calculation method for a mast crane operation end based on BeiDou. The computer program product provided by this invention can solve the technical problems in the prior art, such as poor reliability of a single BeiDou system, lack of inertial navigation drift suppression, lack of kinematic constraint correction for the operation end, and lack of a closed-loop self-optimization mechanism, resulting in low pose calculation accuracy, poor scene adaptability, and insufficient long-term stability. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the multi-source pose fusion calculation method for a mast crane operation end based on BeiDou provided in the above-described embodiment, and will not be repeated here.

[0052] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a read-only memory. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0053] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0054] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A multi-source pose fusion calculation method for a mast crane operating end based on BeiDou, characterized in that, The methods include: Step S10: Synchronously collect BeiDou positioning data, IMU inertial navigation data, winch encoder data and boom tilt angle sensing data of the mast boom. After outlier removal, spatiotemporal alignment and coordinate system preprocessing, output standardized multi-source time series dataset, unified coordinate system transformation parameters and sensor calibration initial values. Step S20: Based on the standardized multi-source time series dataset and sensor calibration initial values ​​obtained in step S10, perform BeiDou carrier phase difference decomposition and strapdown inertial navigation recursive calculation, and output BeiDou high-precision pose, inertial navigation recursive pose, corresponding confidence index and basic kinematic constraint parameters. Step S30: Based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index and sensor calibration initial values ​​output in step S20, multi-source pose initial fusion is performed using adaptive federated Kalman filtering, and the initial fusion pose result, fusion residual sequence and kinematic constraint parameter set are output. Step S40: Based on the initial fusion pose result and kinematic constraint parameter set output in step S30, introduce the execution error correction of the kinematic constraint model at the working end, and output the final pose calculation result and pose accuracy evaluation parameters at the working end. Step S50: Based on the final pose calculation result output in step S40 and the fusion residual sequence output in step S30, perform iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, output the pose calculation steady-state flag and feed back to update the calculation baseline.

2. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 1, characterized in that, In step S10, the steps of simultaneously collecting BeiDou positioning data, IMU inertial navigation data, winch encoder data, and boom tilt angle sensing data from the mast boom, followed by outlier removal, spatiotemporal alignment, and coordinate system preprocessing to obtain a standardized multi-source time-series dataset, unified coordinate system transformation parameters, and initial sensor calibration values, specifically include: Step S101: Collect raw observation data and ephemeris data output by the Beidou dual-mode receiver installed at the top of the boom at a sampling rate of 1Hz. At the same time, collect triaxial acceleration and triaxial angular velocity data output by the IMU installed at the same position at a sampling rate of 200Hz. Collect rope length data output by the winch shaft encoder and pitch angle data output by the boom tilt sensor at a sampling rate of 50Hz. Step S102: Perform multipath suppression and outlier removal on the original BeiDou observation data, perform fourth-order Butterworth low-pass filtering on the IMU data with a cutoff frequency of 20Hz, perform sliding mean denoising on the encoder and tilt data, and interpolate and align all sensor data using the BeiDou second pulse as the time reference to generate an equally spaced standardized multi-source time series dataset. Step S103: Establish the transformation relationship between the BeiDou coordinate system, the inertial navigation coordinate system and the mast crane body coordinate system, calibrate the installation deviation of each sensor and the geometric parameters of the boom, and output the unified coordinate system transformation parameters and the initial calibration values ​​of the sensors.

3. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 2, characterized in that, In step S20, based on the standardized multi-source time-series dataset obtained in step S10, the steps of performing BeiDou carrier phase difference decomposition and strapdown inertial navigation recursive calculation, and outputting BeiDou high-precision pose, inertial navigation recursive pose, and corresponding confidence index specifically include: Step S201: Based on the original BeiDou observation data and the differential data of the base station, the carrier phase differential algorithm is used for calculation. Combined with cycle slip detection and ambiguity fixing, the BeiDou high-precision positioning result and attitude angle at the top of the boom are obtained. At the same time, the BeiDou positioning reliability is calculated based on the number of visible stars and the PDOP value. Step S202: Using the initial sensor calibration value from step S10 as the initial alignment parameter, perform strapdown inertial navigation recursive calculation on the filtered IMU data, output the position, velocity and attitude angle in the inertial navigation system, and calculate the inertial navigation confidence based on the recursive duration and angular velocity noise level. Step S203: Convert the BeiDou pose and inertial navigation pose to the body coordinate system, and output the BeiDou high-precision pose, the inertial navigation recursive pose and their respective confidence indexes. The confidence value ranges from 0 to 1.

4. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 3, characterized in that, In step S201, the BeiDou positioning confidence level is graded according to the number of visible stars and PDOP value. When the number of visible stars is ≥8 and PDOP ≤2, the confidence level is 0.

95. When the number of visible stars is 5~7 and PDOP ≤4, the confidence level is 0.

7. When the number of visible stars is <5, the confidence level is 0.3 or below. The confidence level is >0.

5. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 1, characterized in that, In step S30, based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index, and sensor calibration initial values ​​output in step S20, an adaptive federated Kalman filter is used to perform initial fusion of multi-source poses to obtain the initial fused pose result, fusion residual sequence, and kinematic constraint parameter set. Specifically, this includes: Step S301: Using the initial sensor calibration value as the coordinate system reference, and using BeiDou and inertial navigation as two independent sub-filters, each with its own confidence level as the observation noise weight coefficient, adaptively adjust the observation noise covariance matrix of the sub-filters. Step S302: The federated Kalman filter architecture is adopted. The two sub-filters are updated in parallel. The main filter performs information fusion according to confidence weight to obtain the globally optimal initial fused pose result of the boom top. Step S303: Record the new information sequence of the filter update at each time step as the fusion residual. Simultaneously, after time stamp matching, the basic kinematic constraint parameters are encapsulated into a kinematic constraint parameter set. Output the initial fusion pose result, the fusion residual sequence, and the kinematic constraint parameter set.

6. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 1, characterized in that, In step S40, based on the initial fused pose result and kinematic constraint parameter set output in step S30, the step of introducing the execution error correction of the kinematic constraint model at the working end to obtain the final pose calculation result and pose accuracy evaluation parameters at the working end specifically includes: Step S401: Extract boom tilt angle data and winch rope length data from the kinematic constraint parameter set, combine them with boom geometric parameters, construct a kinematic model of the mast crane working end, and calculate the theoretical pose of the working end spreader from the initial fused pose of the boom top. Step S402: Introduce rope length constraints and swing angle constraints to correct kinematic errors in the initial fusion pose, compensate for pose deviations caused by boom deformation and rope swing, and obtain the three-dimensional position and attitude angle of the lifting device at the working end. Step S403: Calculate pose accuracy evaluation parameters based on the fusion residual and constraint deviation, including horizontal position error estimation and elevation error estimation, and output the final pose calculation result and pose accuracy evaluation parameters of the working end.

7. The multi-source pose fusion calculation method for a mast crane operating end based on BeiDou as described in claim 1, characterized in that, In step S50, based on the final pose calculation result output in step S40 and the fusion residual sequence output in step S30, the steps of performing iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, outputting a pose calculation steady-state flag, and feeding back to update the calculation baseline specifically include: Step S501: When the BeiDou confidence level is higher than the set threshold and the fused residual sequence remains stable, the final pose calculation result is used as the comparison benchmark. The sensor installation deviation and boom geometric parameters are iteratively corrected by combining the high-confidence BeiDou observation values. The optimized calibration parameters are output and fed back to step S10 to update the benchmark. Step S502: Calculate the observation noise bias based on the statistical characteristics of the fused residual sequence, and estimate the process noise level by combining the temporal fluctuation of the final pose solution result. Then, jointly and adaptively adjust the process noise covariance and observation noise weight of the federated Kalman filter to optimize the filter gain of subsequent frames. Step S503: When the fusion residuals of 30 consecutive frames are all less than the set residual threshold, and the pose fluctuation of the final pose calculation result is less than the limit, the calculation is determined to be converged and the pose calculation steady-state flag is set, the steady-state flag is output and the initial baseline of the calculation in step S20 is updated.

8. A BeiDou-based multi-source pose fusion calculation system for a mast crane operating end, applied to the BeiDou-based multi-source pose fusion calculation method for a mast crane operating end as described in any one of claims 1 to 7, characterized in that, include: Multi-source sensor acquisition and preprocessing module: used to simultaneously acquire BeiDou positioning data, IMU inertial navigation data, winch encoder data and boom tilt angle sensor data of mast crane boom. After outlier removal, spatiotemporal alignment and coordinate system preprocessing, it outputs standardized multi-source time series dataset, unified coordinate system transformation parameters and sensor calibration initial values. Dual-source pose independent solution module: Based on the standardized multi-source time series dataset and sensor calibration initial values ​​obtained by the multi-source sensor acquisition preprocessing module, it performs BeiDou carrier phase difference solution and strapdown inertial navigation recursive solution, and outputs BeiDou high-precision pose, inertial navigation recursive pose, corresponding confidence index and kinematic constraint basic parameters. Adaptive filtering initial fusion module: Based on the BeiDou high-precision pose, inertial navigation recursive pose, confidence index and sensor calibration initial value output by the dual-source pose independent solution module, the module uses adaptive federated Kalman filtering to perform multi-source pose initial fusion, and outputs the initial fusion pose result, fusion residual sequence and kinematic constraint parameter set. Kinematics constraint correction module: Based on the initial fusion pose result and kinematic constraint parameter set output by the adaptive filtering initial fusion module, it introduces the kinematic constraint model of the working end to perform error correction, and outputs the final pose solution result and pose accuracy evaluation parameters of the working end; Iterative optimization closed-loop feedback module: Based on the final pose calculation result output by the kinematic constraint correction module and the fusion residual sequence output by the adaptive filtering initial fusion module, it performs iterative optimization of sensor calibration parameters and adaptive adjustment of filter gain, outputs pose calculation steady-state flag and feeds back to update the calculation baseline.

9. A multi-source pose fusion calculation device for a mast crane operating end based on BeiDou, characterized in that, The BeiDou-based mast crane operation end multi-source pose fusion calculation device includes: a memory, a processor, and a BeiDou-based mast crane operation end multi-source pose fusion calculation program stored in the memory and executable on the processor. When the BeiDou-based mast crane operation end multi-source pose fusion calculation program is executed by the processor, it implements the BeiDou-based mast crane operation end multi-source pose fusion calculation method according to any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a BeiDou-based multi-source pose fusion calculation program for mast crane operation terminals. When the BeiDou-based multi-source pose fusion calculation program for mast crane operation terminals is executed by the processor, it implements a BeiDou-based multi-source pose fusion calculation method for mast crane operation terminals according to any one of claims 1 to 7.

Images