Control system and method for steel truss beam spmt moving process

Abstract

The application discloses a control system and method for steel truss beam SPMT moving process, and belongs to the technical field of adaptive control, comprising: a link acquisition module, a judgment module, a regulation and control module and an optimization module; the link acquisition module is used for collecting moving data in real time through the constructed full-dimensional perception network; the judgment module is used for obtaining road condition disturbance parameters, wind load disturbance parameters, and judging the load state of the SPMT moving; the regulation and control module generates pre-action instructions and deviation correction instructions through a rule base and an instruction generator, linearly superimposes the pre-action instructions and the deviation correction instructions according to instruction superposition weights, and generates a control instruction set; the optimization module is used for smoothing the control instruction set to optimize the instruction change curve, and issuing the smoothed control instruction to each SPTM, receiving actual response data after instruction execution, comparing the actual response with the expected response, and updating the rule base; flexible regulation and control under dynamic disturbance are realized.

Description

Technical Field

[0001] This invention relates to a control system and method for the SPMT (Special Purpose Transport Machine) process of steel truss beams, belonging to the field of adaptive control technology. Background Technology

[0002] In the field of bridge engineering construction, the on-site relocation of large-span steel trusses places stringent requirements on equipment and management. Self-propelled modular transporters (SPMTs), with their high load-bearing capacity and modular splicing technology, have become the core equipment for steel truss relocation and have been widely used in the relocation construction of various steel truss projects. The basic parameters of SPMT relocation operations provide basic technical support for steel truss relocation operations, ensuring the smooth progress of relocation construction in conventional scenarios.

[0003] However, existing technologies do not take into account the need for real-time flexible control under dynamic disturbances during the relocation of steel trusses. Specifically, when faced with complex road conditions, wind loads, and other dynamic disturbances during relocation, existing monitoring and control methods cannot generate optimal and flexible adjustment commands in real time. This results in coarse control operations during the relocation of steel trusses, which not only reduces the efficiency of the relocation operation but also risks causing oscillations in the relocation system due to improper control, affecting the safety and stability of the steel truss relocation. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a control system and method for the SPMT (Special Purpose Vehicle) transport process of steel truss girders. By collecting transport data through multi-dimensional sensing, processing the data to obtain disturbance parameters and SPMT load status, generating and superimposing pre-action commands and correction commands, smoothing the commands before issuing them, and updating the rule base in conjunction with response feedback, flexible control under dynamic disturbances can be achieved, solving technical problems such as coarse control and system oscillation.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A control system for the SPMT (Special Purpose Turbine Material Handling) process of steel truss girders includes: a data acquisition module, a decision module, a control module, and an optimization module.

[0007] The link acquisition module is used to collect transportation data in real time, including road elevation data, wind speed and direction data, pressure data, strain data, and vehicle operation data;

[0008] The judgment module obtains road condition disturbance parameters and wind load disturbance parameters through road elevation data and wind speed and direction data. By performing statistical analysis on pressure data and inverse analysis on strain data, it determines the load state of the SPMT being moved.

[0009] The control module is used to call the preset rule base and instruction generator, generate pre-action instructions based on road condition disturbance parameters and wind load disturbance parameters, generate correction instructions by comparing the load state with the preset target state, perform offline global optimization of the rule base through particle swarm optimization algorithm to obtain instruction superposition weight, and linearly superimpose the pre-action instructions and correction instructions according to the instruction superposition weight to generate a control instruction set.

[0010] The optimization module is used to smooth the control instruction set to optimize the instruction change curve, and then send the smoothed control instructions to each SPTM. At the same time, it receives the actual response data after the instruction is executed, compares the actual response with the expected response, and updates the rule base.

[0011] Specifically, the steps for obtaining road condition disturbance parameters include:

[0012] Based on the preset ratio of distance coefficient to real-time vehicle speed, the filter window size is set, and the road elevation data is filtered to obtain smooth elevation data.

[0013] Set an elevation benchmark for an ideal smooth road surface, and calculate the elevation deviation of each sampling point by combining the elevation measurement value of each sampling point in the smooth elevation data.

[0014] Obtain the sampling interval of the sensor array and calculate the spatial coordinate position corresponding to each sampling point;

[0015] By associating the elevation deviation of each sampling point with its corresponding spatial coordinates, a road surface excitation map is constructed.

[0016] Based on the physical parameters and second-order closed-loop transfer function of the SPMT suspension terminal, the vibration transmission signal of the transfer terminal is calculated.

[0017] The elevation deviation is subjected to Laplace transform to obtain the elevation frequency domain signal. The vertical acceleration, including the front axle acceleration and the rear axle acceleration, is obtained by performing an inverse Laplace transform on the product of the vibration transmission signal and the elevation frequency domain signal.

[0018] By calculating the acceleration difference between the front and rear axles, the equivalent mass, and the wheelbase, the pitching moment is calculated. The road excitation map, pitching moment, and vertical acceleration are then integrated to form road condition disturbance parameters.

[0019] Specifically, the steps for obtaining wind load disturbance parameters include:

[0020] Smooth the wind speed and direction data and verify the data validity to generate smoothed wind data;

[0021] Retrieve the wind load factor lookup table to find the benchmark wind load factor corresponding to the current wind speed;

[0022] Based on the wind direction angle corresponding to the current wind speed, the corresponding wind direction correction factor is selected from the pre-built correction rule table;

[0023] If no wind direction correction factor is found, select two angles adjacent to the current wind direction angle and their corresponding wind direction correction factors, and obtain the wind direction correction factor for the current wind direction angle through linear interpolation.

[0024] Based on the wind direction correction factor and the baseline wind load factor, the corrected baseline coefficient is obtained, and then combined with the wind direction angle to be decomposed into the lateral force wind load factor and the lift wind load factor.

[0025] Obtain the windward area of ​​the steel truss and, in conjunction with the formula for calculating air resistance, calculate the static equivalent wind load, including lateral force wind load and lift wind load.

[0026] The distance between the point of application of the wind load and the geometric center of the steel truss is obtained. The torsional moment is calculated by combining the static equivalent wind load and the static equivalent wind load, and the wind load disturbance coefficient is formed.

[0027] Specifically, determining the load status of SPMT migration includes:

[0028] Based on the pressure data, the mean, variance, and range of the pressure at each point are calculated. The eccentric load coefficient is obtained by the ratio of the pressure range to the mean, and the pressure distribution characteristics are obtained.

[0029] The strain data were processed based on Hooke's law to obtain the stress values ​​of key sections of the steel truss beam.

[0030] Based on the geometric parameters of the steel truss section and the bonding design of strain gauges, the stress state of key sections of the steel truss is determined, including tensile stress and compressive stress.

[0031] Based on vehicle operation data, calculate the attitude offset and attitude change rate during the transfer process;

[0032] By integrating pressure distribution characteristics, wind load disturbance parameters, and additional forces generated by attitude deviation, the internal force parameters of key sections of the steel truss girder, including axial force, shear force, and bending moment, are calculated based on the static equilibrium equation.

[0033] Correlation analysis was used to calculate the influence weight of each disturbance parameter on the internal force parameter, and a collaborative correlation map was constructed by associating road condition disturbance parameters, wind load disturbance parameters, and internal force parameters.

[0034] Specifically, determining the load status of SPMT migration also includes:

[0035] Obtain state determination thresholds based on offline experimental calibration, including off-center load threshold, pressure range threshold, stress threshold, and attitude deviation threshold;

[0036] Based on the off-center load factor, pressure difference, beam stress value, and attitude offset, the current load state is determined, including load balance state, slight off-center load state, and severe off-center load state.

[0037] Based on historical data under all transport conditions, instability trend characteristics are extracted, including pressure difference growth rate, stress change rate, and attitude deviation acceleration.

[0038] Obtain a preset trend threshold and combine it with the characteristics of the instability trend to determine whether there is a potential risk of instability;

[0039] If any unstable trend feature continues to exceed the trend threshold within 3 consecutive sampling points, it is determined that there is a potential risk of instability, and the risk level is marked accordingly.

[0040] Based on the assessment of the current load status and potential instability risks, a judgment report is generated for the current SPMT migration.

[0041] Specifically, the rule base stores road disturbance feedforward rules, wind load disturbance feedforward rules, load deviation feedback rules, and attitude deviation feedback rules. Each rule has preset trigger conditions, execution actions, and associated parameters, including:

[0042] The triggering condition for the road disturbance feedforward rule is that the elevation deviation is greater than the local bump threshold and the real-time vehicle speed is greater than the vehicle speed threshold. The execution action is to generate a preset speed curve command to smoothly decelerate to the preset safe vehicle speed.

[0043] The triggering condition for the wind load disturbance feedforward rule is that the lateral wind load is greater than the wind load threshold and the SPMT is traveling in a straight line. The execution action is to generate a preventive load compensation command to increase the adjustment differential pressure.

[0044] The load deviation feedback rule is triggered when the pressure range is continuously greater than the pressure range threshold, and the action is to generate a command to enable the pressure PID feedback control algorithm.

[0045] The attitude deviation feedback rule is triggered when the attitude deviation is continuously greater than the attitude deviation threshold, and the action is to generate a command to enable the attitude PID feedback control algorithm.

[0046] Specifically, the steps for generating a control instruction set include:

[0047] The system calls up road condition disturbance parameters and wind load disturbance parameters, compares them with the rule base to match the corresponding feedforward rules, and generates pre-action instructions.

[0048] Based on the safety threshold for feedforward instructions in the rule base, the validity of pre-action instructions is verified to eliminate invalid instructions, and the triggering rules and corresponding disturbances of each instruction are marked.

[0049] The pressure range and attitude offset are called, and the rule base is compared to match the corresponding feedback rule. The corresponding controller is called to generate correction instructions, including pressure correction instructions and attitude correction instructions.

[0050] The instruction weights in the rule base are retrieved, and the same type of instructions in the pre-action instructions and correction instructions are linearly superimposed to calculate the adjustment amount, so as to generate the initial control instructions.

[0051] Specifically, the steps for generating a control instruction set include:

[0052] Obtain an offline working condition dataset containing instruction superposition weights, use the instruction superposition weights as candidate values ​​for particle positions, and construct an objective function based on the transport evaluation index;

[0053] The working condition parameters in the offline working condition dataset are traversed, the mean of the objective function for all working conditions under different instruction superposition weights is calculated, and the particle position is iteratively updated with the goal of maximizing the mean of the objective function. The optimal instruction superposition weight is obtained and written into the rule base.

[0054] For the initial control command, commands whose adjustment amount is greater than the command superposition threshold are removed, and the removed commands are replaced with the command superposition threshold as the corresponding adjustment amount.

[0055] Based on the vehicle speed command and road excitation map in the initial control command, linear interpolation is used to decompose the vehicle speed command into the target vehicle speed value of each road segment and generate the target speed curve.

[0056] Based on the pressure command in the initial control command and the pressure distribution characteristics of each point, the pressure command is broken down into the target pressure value of each point.

[0057] Based on the steering angle command, attitude offset, and sampling period in the initial control command, the steering angle command is decomposed into the target steering angle value for each sampling period to form a target steering angle sequence;

[0058] The target speed curve, target pressure value, and target steering angle sequence are integrated into a control command set, and the execution conditions and execution priority of each command are marked.

[0059] Specifically, the steps to optimize the instruction variation curve include:

[0060] The system retrieves preset speed smoothing thresholds, pressure change thresholds, and steering fine-tuning limits. Based on the control instruction set, it identifies instruction abrupt change points, including speed abrupt change points, pressure abrupt change points, and steering abrupt change points.

[0061] For speed change points, instruction segments with one sampling period are set before and after the speed change point to form a transition segment, and a continuous transition speed is generated by linear interpolation.

[0062] For pressure abrupt change points, the exponential smoothing coefficient is called to determine the direction of pressure adjustment based on the pressure adjustment amount, and the pressure adjustment amount is increased or decreased in combination with the exponential smoothing coefficient.

[0063] For a sudden steering change point, the excess adjustment amount is obtained by the difference between the steering angle adjustment amount and the upper limit of the steering fine adjustment. A fine adjustment cycle is added after the original sudden steering change point to form a steering transition segment.

[0064] Using the sine term, the cumulative fine-tuning ratio for each fine-tuning cycle is calculated. The single fine-tuning ratio for the current fine-tuning cycle is obtained by the difference between the cumulative fine-tuning ratios of adjacent fine-tuning cycles. Combined with the overshoot adjustment, the single fine-tuning amount for the current fine-tuning cycle is obtained.

[0065] The smoothed control instructions are then used to replace the segments at the abrupt change points in the original control instruction set, thus forming smoothed control instructions.

[0066] Control methods for the SPMT (Special Purpose Material Handling) process of steel truss girders include:

[0067] By utilizing the constructed full-dimensional perception network, the transportation data of the entire process of moving steel truss SPMT is collected in real time and preprocessed.

[0068] Based on the transport data, disturbance parameters are obtained, and the load status of the SPMT transport is determined. The disturbance parameters include road condition disturbance parameters and wind load disturbance parameters.

[0069] The system calls a preset rule base, triggers feedforward rules with disturbance parameters to generate pre-action instructions, and triggers feedback rules with load status to generate correction instructions. The system constructs a control instruction set by linearly superimposing these instructions.

[0070] The control command set is smoothed to optimize the command change curve and then sent to each SPTM local controller for updating based on actual response data.

[0071] The beneficial effects of this invention are:

[0072] By constructing a multi-dimensional perception network, various types of migration data are collected in real time to comprehensively capture dynamic disturbances and system operating status during the migration process. This allows for the acquisition of disturbance parameters and determination of SPMT load status, providing comprehensive and real-time basic data support for subsequent instruction generation. This avoids blind instruction generation due to data gaps. Furthermore, a preset rule base and instruction generator are invoked, and pre-action instructions are generated based on disturbance parameters to proactively address the impact of disturbances. Corrective instructions are generated by comparing load status with target status to correct operational deviations. Finally, a control instruction set is generated by linearly superimposing two types of instructions with weighted superposition, achieving flexible and precise adaptation of instructions. This allows for the real-time generation of optimal adjustment instructions, avoiding migration errors caused by coarse control operations. To address the issue of low transport efficiency, this system reduces the risk of system oscillations caused by poor command adaptability. It smooths the control command set, optimizes command change curves, avoids SPMT (Special Power Management) operational shocks caused by sudden command changes, and ensures stable execution after command issuance. Simultaneously, it receives actual response data and compares it with expected responses, dynamically updating the rule base to continuously optimize the control strategy. This further enhances the system's adaptability to complex road conditions and wind load disturbances, eliminating oscillation problems caused by improper control. The overall closed-loop control not only improves the efficiency of steel truss girder relocation operations but also significantly enhances the safety and stability of the relocation system. It achieves real-time, flexible, and precise control under dynamic disturbances, adapting to various relocation conditions without complex operations. Attached Figure Description

[0073] Figure 1 This is a structural diagram of the control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process for steel truss girders.

[0074] Figure 2 This is a flowchart illustrating the process of determining the load state during SPMT transport in this invention.

[0075] Figure 3 This is a flowchart illustrating the generation of the control instruction set in this invention;

[0076] Figure 4 This is a flowchart of a control method for the SPMT (Special Purpose Turbine Material Handling) process for steel truss beams. Detailed Implementation

[0077] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0078] Example 1:

[0079] refer to Figures 1 to 3As shown, this embodiment introduces a control system for the SPMT (Special Power Module) transport process of steel truss girders, including: a link acquisition module, a judgment module, a control module, and an optimization module;

[0080] The link acquisition module is used to build a full-dimensional perception network by deploying multi-source sensors, and to build a low-latency, high-reliability wireless transmission link by combining a wireless transmission network. It collects and preprocesses the transportation data of the entire process of moving steel truss SPMT in real time, including road elevation data, wind speed and direction data, pressure data, strain data, and vehicle operation data.

[0081] The sensor deployment locations include preset points in the SPMT and key sections of the main steel truss structure. The sensors include an array of optical / ultrasonic ranging sensors, mechanical anemometers, high-precision pressure transmitters, resistance strain gauges, dual-axis tilt sensors, and reflecting prisms. Road elevation data is acquired through the optical / ultrasonic ranging sensor array to capture the terrain undulations of the path ahead. Wind speed and direction data are collected by mechanical anemometers installed on the top and four corners of the steel truss to provide feedback on real-time environmental wind conditions. Pressure data is obtained through high-precision pressure transmitters integrated into the hydraulic suspension circuits of each SPMT, reflecting the working pressure within the SPMT hydraulic suspension circuits and indirectly indicating the load distribution. Strain data is collected through resistance strain gauges attached to key sections of the main steel truss structure, presenting the degree of minute deformation of the steel truss under stress in the form of electrical signals. Vehicle operation data includes attitude, position, speed, and steering angle.

[0082] The judgment module is used to perform moving average filtering on road elevation data to obtain ideal elevation deviation. Combined with current vehicle operation data, it generates road excitation maps of the current and forward road sections to obtain road condition disturbance parameters in the future period. It also uses wind speed and direction data and a constructed wind load coefficient lookup table to obtain wind load disturbance parameters in real time. It performs statistical analysis on pressure data to obtain pressure distribution at each SPMT preset point. Combined with the structural stress changes reflected by strain data and the attitude offset reflected by vehicle operation data, it determines the load state of SPMT transfer to identify unbalanced load state and potential instability trend in real time during SPMT transfer.

[0083] The control module is used to call the preset rule base and instruction generator. Based on road condition disturbance parameters and wind load disturbance parameters, it triggers the feedforward rules in the rule base to generate pre-action instructions to offset the impact of disturbances in advance. Based on the comparison between the load state and the preset target state, it triggers the feedback rules in the rule base to generate correction instructions. The rule base is then optimized offline globally using the particle swarm optimization algorithm to obtain the instruction superposition weight. Based on the instruction superposition weight, the pre-action instructions and correction instructions are linearly superimposed to generate a control instruction set that includes the target speed curve, pressure setpoint, and steering angle sequence, ensuring that the instructions are flexible and easy to execute.

[0084] The optimization module smooths the control command set to optimize the command change curve and avoid SPMT operation shocks caused by sudden command changes. It sends the smoothed control commands to each SPTM local controller through a secure communication protocol, and receives the actual response data after command execution. It compares the actual response with the expected response, calculates the absolute response deviation, and iteratively calculates the correction increment value of each parameter in the rule base through the PID controller. Based on the correction increment value, it performs targeted iterative updates to the feedforward control rules, feedback control rules, and command superposition weight rules in the rule base to update the rule base. This ensures operational stability through secure command delivery and continuously improves the system's adaptability to complex road conditions and wind load disturbances through feedback adjustments.

[0085] Furthermore, the steps for obtaining road condition disturbance parameters and wind load disturbance parameters include:

[0086] Obtain the distance coefficient determined by those skilled in the art based on the engineering calibration, such as setting it to 2m, calculate the ratio of the distance coefficient to the real-time speed of the SPMT, round this ratio, obtain the size of the filtering window, and filter the road elevation data to obtain smooth elevation data.

[0087] Based on the design path elevation surveyed before the relocation operation, an elevation benchmark for an ideal flat road surface is set.

[0088] For each sampling point in the smoothed elevation data, the elevation deviation of each sampling point is calculated by the difference between the elevation measurement value and the elevation benchmark. By integrating the elevation deviations of all sampling points, a deviation set is generated.

[0089] The sampling interval of the sensor array is obtained. The origin of the coordinate system is taken as the front end of the SPMT, and the x-axis coordinate system is established only along the path travel direction. Since the sensor array is installed along the path direction of the SPMT end, the collection range is focused on the path directly in front of the movement. There is no need to consider the y-axis parallel to the road surface. The spatial coordinate position corresponding to each sampling point is calculated through the sampling interval.

[0090] The elevation deviation of each sampling point is associated with the corresponding spatial coordinates. With the spatial position as the horizontal axis and the elevation deviation as the vertical axis, a road excitation map of the current and forward preset lengths is constructed. The preset length is determined by those skilled in the art based on the SPMT vehicle parameters, transportation conditions, and control response characteristics, and is set to 10m in this embodiment.

[0091] For accelerations directly acquired from the front and rear axles, there is a significant amount of invalid interference. Furthermore, the directly acquired accelerations represent the vibration response already generated by the front and rear axles at the current moment, making them post-hoc data. Using the physical parameters of the SPMT suspension terminal, such as hydraulic suspension stiffness, damping coefficient, and equivalent mass, the angular frequency and damping ratio of the transfer terminal are calculated based on the closed-loop transfer function of the second-order system. Calculate the vibration transmission signal of the transfer terminal; among which, The Laplace operator is used to convert the time-domain signal to the frequency-domain signal. The suspension terminal includes the front suspension axle and the rear suspension axle. Therefore, the vibration transmission signal calculated here includes the front axle vibration transmission signal and the rear axle vibration transmission signal.

[0092] The elevation deviation is subjected to Laplace transform to obtain the elevation frequency domain signal. The frequency domain result of the vertical vibration response is obtained by multiplying the vibration transmission signal and the elevation frequency domain signal. The inverse Laplace transform of this frequency domain result is then performed to obtain the vertical acceleration in the time domain, including the front axle acceleration and the rear axle acceleration.

[0093] The acceleration difference between the front and rear axles is calculated. The pitching moment is obtained by multiplying the equivalent mass, the wheelbase between the front and rear axles, and the acceleration difference. The road excitation map, pitching moment, and vertical acceleration are integrated to form quantified road condition disturbance parameters.

[0094] For example, smoothed elevation data is When the elevation datum is 0 and the sampling interval is 0.5m, the resulting deviation set is: With the SPMT front end as the origin and the x-axis as the driving direction, the spatial coordinates of each sampling point are 0.5m, 1m, 1.5m, 2m, and 2.5m. With x as the horizontal axis and elevation deviation as the vertical axis, the correlation result is... , , , , This forms the road surface excitation curve within 2.5m directly in front of the transport, i.e., the road surface excitation map;

[0095] The physical parameters of the SPMT suspension terminal include hydraulic suspension stiffness. Damping coefficient Equivalent quality wheelbase between front and rear axles The parameters of a second-order system include angular frequency. Damping ratio Closed-loop transfer function ;

[0096] Perform a Laplace transform on the elevation deviations in the deviation set, such as the sampling points. Convert to ,and Multiplication yields the frequency domain result. The inverse transform yields the time-domain vertical acceleration, including the front-axis acceleration. Rear axle acceleration ;

[0097] At this time, the acceleration difference pitching moment ;

[0098] Set the smoothing window duration, such as 1 second, and use the mean of the smoothing window to smooth the wind speed and direction data. At the same time, perform data validity verification. Only if the wind speed data of three consecutive sampling points exceeds the effective measurement range of the sensor at least twice or the wind direction angle is greater than the full angle, it is marked as invalid data and the mean of the previous valid window is used to fill in the data, and finally the smoothed wind force data is generated.

[0099] The wind load coefficient lookup table constructed from wind tunnel tests was retrieved to find the reference wind load coefficient corresponding to the current wind speed. The initial wind direction angle corresponding to the reference wind load coefficient is 0°, indicating that the wind direction is parallel to the axis of the steel truss. The wind load coefficient lookup table was pre-set by those skilled in the art based on the wind-resistant design specifications for highway bridges and the actual measurement data from wind tunnel tests. The 0° wind direction angle (wind direction parallel to the axis of the steel truss) was set as the reference working condition. In the commonly used wind speed range of 0~25m / s for the movement of steel trusses, the wind resistance coefficient of the steel truss was tested at different wind speeds with a step size of 1m / s. After verification by the technicians, it was used as the reference wind load coefficient for the corresponding wind speed.

[0100] Since the wind direction angle at the current wind speed is not 0°, the offline calibration correction rule table is invoked. Based on the wind direction angle corresponding to the current wind speed, the corresponding wind direction correction factor is filtered from the correction rule table. If no matching wind direction correction factor exists in the correction rule table, two angles adjacent to the current wind direction angle and their corresponding wind direction correction factors are selected from the correction rule table. The wind direction correction factor for the current wind direction angle is obtained through linear interpolation. For example, if the wind direction angle is 35°, but there is no corresponding wind direction correction factor for 35° in the table, then the wind direction correction factors for the adjacent angles of 0° and 90° are selected. , Thus, the wind direction correction factor corresponding to 35° is calculated. The correction rule table was constructed by those skilled in the art based on wind tunnel test data from the same batch.

[0101] The corrected reference coefficient is obtained by multiplying the wind direction correction factor and the reference wind load coefficient. The corrected reference coefficient is then decomposed into the lateral force wind load coefficient and the lift wind load coefficient by combining the wind direction angle.

[0102] Based on the cross-sectional dimensions of the steel truss, the windward area of ​​the steel truss is obtained. Using the formula for calculating air resistance, and taking the wind load factor as the drag factor, along with air density, wind speed, and windward area, the static equivalent wind load acting on the steel truss is calculated, including lateral wind load and lift wind load. Where the windward area... air density Actual wind speed When the wind direction angle is 35°, the split lateral force wind load coefficient Lift wind load factor Then the corresponding lateral wind load can be calculated. Lift wind load ;

[0103] The distance between the wind load application point and the center of the steel truss girder is obtained, including the horizontal distance between the lateral force application point and the center of the girder, and the vertical distance between the lift application point and the center of the girder. The corresponding static equivalent wind load is then multiplied, and the product values ​​of each force are superimposed to obtain the torsional moment. Combined with the static equivalent wind load, the wind load disturbance coefficient is formed. The wind load application point is the location of the resultant wind load force measured in the wind tunnel test.

[0104] Furthermore, determining the load status of the SPMT during transport includes:

[0105] Based on the pressure data, the mean, variance, and range of the pressure at each point are calculated. The eccentric load coefficient is obtained by the ratio of the pressure range to the mean, so as to obtain the pressure distribution characteristics.

[0106] The strain data were processed based on Hooke's law to obtain the stress values ​​of key sections of the steel truss beam.

[0107] Based on the geometric parameters of the steel truss section and the bonding design of strain gauges, the stress state of the key sections of the steel truss is determined, including tensile stress and compressive stress. The bonding design of the strain gauges involves pre-setting a bonding scheme on the key sections of the main structure of the steel truss. Resistance strain gauges are bonded to the tensile stress zone (such as the tension zone at the bottom of the beam) and the compressive stress zone (such as the compression zone at the top of the beam) of the key sections of the steel truss, and the bonding area of ​​each strain gauge is marked in advance.

[0108] Based on vehicle operation data, the attitude offset and attitude change rate during the transfer process are calculated. The attitude offset includes tilt angle deviation and position offset. The tilt angle includes pitch angle and roll angle. The tilt angle deviation is the difference between the tilt angle data and the theoretical attitude tilt angle. The position offset is the difference between the positioning data and the preset design path coordinates. The attitude change rate is the first derivative of the tilt angle deviation and position offset with respect to time, reflecting the rate of change of attitude offset, such as the roll angle change rate and the lateral position offset change rate.

[0109] Taking the key sections of the steel truss as the research object, this study integrates pressure distribution characteristics, wind load disturbance parameters, and additional forces generated by attitude deviation. Based on the static equilibrium equations (horizontal force equilibrium, vertical force equilibrium, and moment equilibrium), the internal force parameters of the key sections of the steel truss are inverted, including axial force, shear force, and bending moment. The inversion process includes: in the horizontal force equilibrium, the sum of the lateral wind load and the horizontal additional force of attitude deviation is calculated. Since the sum of the resultant force of all external loads and the axial force of the section is 0, the negative value of this sum is taken as the axial force. In the vertical force equilibrium, the sum of the lift wind load, the vertical additional force of attitude deviation, and the resultant force of all SPMT support forces on the isolator is calculated. The shear force is obtained by the difference between the self-weight of the isolator and this sum. In the moment equilibrium, based on the fact that moment is the product of the magnitude of the force and the lever arm, the total moment of the SPMT support force, the total moment of the wind load, the total moment of the attitude deviation additional force, and the moment of the self-weight of the isolator are calculated. The bending moment is obtained by the negative value of the sum of the moments.

[0110] For example, if the externally acquired attitude offset horizontal additional force is 1.2kN, and the calculated lateral wind load is -5.07kN, then the corresponding axial force is... The negative sign in -5.07kN indicates that the lateral wind load is in the negative direction of the x-axis.

[0111] The isolator mass is 50,000 kg, corresponding to an isolator self-weight of 490 kN. The externally acquired SPMT support force is 290 kN, the vertical additional force due to attitude deviation is -0.8 kN, and the calculated lift wind load is 7.26 kN. Therefore, the corresponding shear force is... ;

[0112] Since the SPMT support force has two points of application, with coordinates of -30m and -15m respectively, corresponding to support forces of 150kN and 140kN, the total torque of the SPMT support force is: The coordinates of the point of application of the wind load are -10m horizontally and 2m perpendicular to the ground. Combining the lateral wind load and the lift wind load, the total moment of the wind load is... The coordinates of the point of application of the attitude offset additional force are -12m in the horizontal direction and 1.5m in the vertical direction. Therefore, the total torque of the attitude offset additional force is... If the coordinates of the point of application of the self-weight are -25m, then the torque of the self-weight of the isolated body is... The final bending moment is ;

[0113] Correlation analysis was used to calculate the influence weight of each disturbance parameter on the internal force parameter, and a collaborative correlation map was constructed to present the correlation between the factors.

[0114] The system acquires state determination thresholds based on offline experimental calibration, including off-center load threshold, pressure range threshold, stress threshold, and attitude deviation threshold. It then links these thresholds with pressure distribution characteristics (off-center load coefficient, pressure range), beam stress value, and attitude deviation to determine the current load state, including balanced load state, slight off-center load state, and severe off-center load state. During the determination process, the system prioritizes using collaborative correlation maps to exclude parameter fluctuations caused by normal disturbances. Specifically, a balanced load state is defined as where the off-center load coefficient, pressure range, stress value, and attitude deviation are all below the state determination thresholds; a slight off-center load state is defined as where only one parameter exceeds the state determination threshold, while the remaining parameters are below their corresponding thresholds; and a severe off-center load state is defined as where at least two types of parameters exceed the state determination thresholds. These thresholds are calibrated by those skilled in the art based on statistical analysis of offline experimental data, such as setting the off-center load threshold to 0.15, the pressure range threshold to 0.3 MPa, the stress threshold to 150 MPa, and the attitude deviation threshold to 0.5°.

[0115] Based on historical data under all transport conditions, instability trend characteristics are extracted, including pressure difference growth rate, stress change rate, and attitude deviation acceleration. Among them, pressure difference growth rate is the change in pressure difference per unit time, stress change rate is the instantaneous change rate of stress in key sections, and attitude deviation acceleration is the second derivative of attitude deviation with respect to time.

[0116] Preset trend thresholds are obtained, including general trend thresholds and emergency trend thresholds. These are combined with instability trend characteristics to determine whether there is a potential instability risk. If any instability trend characteristic exceeds the trend threshold for three consecutive sampling points, a potential instability risk is identified, and the risk level is marked. If any instability trend characteristic is greater than the emergency trend threshold, it is marked as high risk. If all instability trend characteristics are between the general trend threshold and the emergency trend threshold, it is marked as low risk. The trend thresholds are constructed through statistical analysis of the full-process historical data of similar steel truss SPMT transport projects of the same type and scale. For example, the general trend threshold for the pressure difference growth rate is 0.05 MPa / s, and the emergency trend threshold is 0.1 MPa / s.

[0117] Based on the assessment of the current load status and potential instability risks, a judgment report is generated for the current SPMT migration.

[0118] Furthermore, the rule base stores control strategies summarized from domain expert experience and offline simulation analysis, including road disturbance feedforward rules, wind load disturbance feedforward rules, load deviation feedback rules, and attitude deviation feedback rules. All rules have preset trigger conditions, execution actions, and associated parameters. The specific basic rules are as follows:

[0119] The triggering condition for the road disturbance feedforward rule is that the elevation deviation is greater than the local bump threshold and the real-time vehicle speed is greater than the vehicle speed threshold. The execution action is to generate a preset speed curve command to smoothly decelerate to a gentle speed. The associated parameters are the local bump threshold, the vehicle speed threshold, and the gentle speed, all of which are the basic values ​​calibrated by offline simulation.

[0120] The triggering condition for the wind load disturbance feedforward rule is that the lateral wind load is greater than the wind load threshold and the SPMT is traveling in a straight line. The execution action is to generate a preventive load compensation command to increase the differential pressure adjustment value of the upwind SPMT group. The associated parameters are the wind load threshold and the differential pressure adjustment, both of which are the basic values ​​calibrated by offline simulation. Among them, traveling in a straight line means that the SPMT does not make a turning operation at the traffic light.

[0121] The load deviation feedback rule is triggered when the pressure difference is continuously greater than the pressure difference threshold. The action is to generate a command to enable the pressure PID feedback control algorithm to dynamically fine-tune the suspension height of the pressure too high / too low groups so that the pressure tends to be balanced. The associated parameters are the pressure difference threshold, the pressure duration range, and the pressure PID controller parameters, which are obtained by offline calibration. The continuous determination criterion is that the pressure difference is greater than the pressure difference threshold within the preset pressure duration range.

[0122] The trigger condition for the attitude deviation feedback rule is that the attitude deviation is continuously greater than the attitude threshold. The action is to generate a command to enable the attitude PID feedback control algorithm, so as to fine-tune the SPMT steering angle and the corresponding preset point suspension height, correct the attitude deviation, and avoid the beam tilting. The associated parameters are the attitude threshold, the attitude duration range, and the attitude PID controller parameters, all of which are calibrated offline through simulation.

[0123] Furthermore, the steps for generating a control instruction set include:

[0124] The road condition disturbance parameters and wind load disturbance parameters generated by the judgment module are called, and the corresponding feedforward rules are matched with the rule library. Pre-action instructions are generated. If the road disturbance feedforward rules and the wind load disturbance feedforward rules are triggered at the same time, the speed adjustment and pressure adjustment instructions are superimposed.

[0125] The validity of pre-action commands is verified by comparing them with the safety thresholds for feedforward commands in the rule base, such as the smooth speed not being lower than the minimum safe speed of SPMT and the adjustment of differential pressure not exceeding the safe range of suspension pressure, so as to eliminate invalid commands and mark the triggering rules and corresponding disturbances for each command.

[0126] The system calls upon the pressure range and attitude offset in the judgment module, compares them with the rule base to match the corresponding feedback rules, and calls the given classic controller. For load deviation, aiming to balance the pressure at each preset point, it dynamically fine-tunes the suspension height of the pressure-over-high / under-high groups and calculates the corresponding pressure correction command. For attitude deviation, it calculates the steering angle correction and suspension height fine-tuning based on the difference between the attitude offset and the attitude threshold, generating the attitude correction command. Finally, it generates the correction command. The membership function and rule table of the PID controller have been pre-optimized and set offline. The expression for the correction amount in the correction command is as follows:

[0127]

[0128]

[0129]

[0130] In the formula, The pressure correction amount for the preset points of SPMT. The PID proportional coefficient for pressure correction is pre-optimized and set offline to control the response speed of pressure correction. For extremely poor pressure, The differential pressure threshold. This represents the steering angle correction amount for a single sampling period. This is the PID proportional coefficient for steering angle correction. This is the actual position offset. For location threshold, This represents the fine-tuning amount of the suspension height for a single sampling period. This is the PID proportional coefficient for suspension height correction. This represents the actual tilt angle deviation. The tilt angle threshold, This is the SPMT point spacing coefficient. , , Based on the PID parameter tuning specifications and considering the response characteristics of the SPMT hydraulic system and steering system, initial values ​​for the proportional coefficient are given, such as... Initial values ​​are 0.5~1.0. Initial values: 0.7~1.2 With initial values ​​of 0.4 to 0.8, a dynamic simulation model of SPMT movement was built to simulate different load deviations and attitude deviations. The proportional coefficient was iteratively adjusted using the Ziegler-Nichols tuning method to ensure that the overshoot of the correction response was less than 5% and the settling time was less than 0.5s. The pressure threshold, position threshold, and tilt angle threshold were all calibrated by those skilled in the art based on the path requirements of the steel truss girder movement and the movement capacity of the SPMT.

[0131] By collecting offline simulation data and field test calibration data, an offline working condition dataset covering all transportation conditions is constructed. Each data set includes the corresponding transportation data and labeled command weights. The dataset contains all typical working conditions, such as normal road conditions, extreme wind load, slight off-center load, and severe off-center load.

[0132] The particle position is determined by superimposing command weights on candidate values, and an objective function is constructed based on the transport evaluation index. The transport evaluation index includes load balance and attitude stability. The objective function is obtained by weighting the transport evaluation index. The weights are set by those skilled in the art. Load balance is used to characterize the uniformity of hydraulic pressure distribution at each support point of the SPMT, and is the difference between 1 and the off-center load coefficient. Attitude stability is used to characterize the control effect of attitude deviation during the transport of the steel truss. The ratio of actual tilt angle deviation to tilt angle threshold and the ratio of actual position deviation to position deviation are calculated. The arithmetic mean of these two ratios is calculated, and the attitude stability is obtained by the difference between 1 and the arithmetic mean.

[0133] Traverse the transport data in the offline working condition dataset, for each group of instructions, superimpose weight candidate values, retrieve the transport data of the whole working condition and calculate the corresponding transport evaluation index, and take the mean of the objective function as the comprehensive control effect of the current group of instructions superimposed with weight candidate values ​​under the whole working condition.

[0134] With the goal of maximizing the mean of the objective function as the optimization direction, the particle position is iteratively updated to update the candidate values ​​of the instruction superposition weight. The optimal instruction superposition weight is obtained by finally converging the particle position, i.e., under the full working condition covered by the original data, and written into the rule base.

[0135] For the initial control command, the command superposition threshold is invoked to remove commands whose adjustment amount is greater than the command superposition threshold. For example, if the adjustment amount of the vehicle speed is greater than the command superposition threshold of the vehicle speed, the superposition weight and source of each command are marked. For the removed command, the corresponding adjustment amount is based on the command superposition threshold to replace the removed command.

[0136] Based on the vehicle speed command and road excitation map in the initial control instructions, linear interpolation is used to decompose the vehicle speed command into target vehicle speed values ​​for each road segment, thereby generating a continuous and smooth target speed curve. This clarifies the timing and magnitude of vehicle speed switching for different road segments, ensuring a smooth speed transition and offsetting road disturbances. For example, if the current vehicle speed is 2 km / h and the target vehicle speed in the initial control command set is 0.5 km / h, the road excitation map is divided into three road segments with 0.5m and 1.5m as boundary points: segment 1, segment 2, and segment 3. The target vehicle speeds for each road segment are 2 km / h, 1 km / h, and 0.5 km / h, respectively, ultimately forming a continuous and smooth target speed curve 2 km / h → 1 km / h → 0.5 km / h.

[0137] Based on the pressure command in the initial control command and the pressure distribution characteristics of each SPMT point, the pressure command is broken down into the target pressure value of each SPMT point, and the pressure adjustment sequence and adjustment range of each point are clarified to ensure balanced load distribution and offset wind load disturbance and load deviation. Among them, the suspension height adjustment and pressure adjustment are carried out simultaneously to merge the suspension height adjustment into the pressure adjustment.

[0138] Based on the steering angle command, attitude offset, and sampling period in the initial control command, the steering angle command is decomposed into the target steering angle value for each sampling period, forming a continuous target steering angle sequence. The steering angle adjustment amount for each sampling period is clearly defined to ensure that the SPMT moves along the preset path, corrects attitude offset, and avoids beam tilting.

[0139] The target speed curve, target pressure value, and target steering angle sequence are integrated into a complete synthetic control instruction set. The execution conditions (such as road segment location, disturbance level, and status level) and execution priority of each instruction are marked. At the same time, the basis for the generation of the instruction (the triggering rule and the corresponding disturbance / deviation) is marked to ensure that the instruction can be directly issued to each SPMT local controller for execution. Among them, the execution priority is based on the rule setting, with pressure correction instructions taking precedence over speed adjustment instructions.

[0140] Furthermore, the steps to optimize the instruction variation curve include:

[0141] The system retrieves preset vehicle speed smoothing thresholds, pressure change thresholds, and steering fine-tuning upper limits. Based on the control command set, it identifies command abrupt change points, including speed abrupt change points, pressure abrupt change points, and steering abrupt change points. Specifically, a speed abrupt change point occurs when the acceleration within the sampling period exceeds the vehicle speed smoothing threshold; a pressure abrupt change point occurs when the absolute value of the pressure adjustment within the sampling period exceeds the pressure change threshold; and a steering abrupt change point occurs when the steering angle adjustment within the sampling period exceeds the steering fine-tuning upper limit. The vehicle speed smoothing threshold, pressure change threshold, and steering fine-tuning upper limit are calibrated by those skilled in the art using standardized procedures based on characteristic analysis, offline simulation, and engineering testing. For example, in this embodiment, the vehicle speed smoothing threshold is set to 2 km / h. 2 The pressure change threshold is 0.2 MPa / sampling cycle, and the upper limit for steering fine-tuning is 3° / sampling cycle;

[0142] For each identified instruction abrupt change point, a smooth transition optimization is performed.

[0143] For speed change points, instruction segments with one sampling period are set before and after the speed change points to form transition segments. The starting point of the transition segment is the current vehicle speed and the ending point is the target vehicle speed. Continuous transition speeds are generated by linear interpolation to ensure that the vehicle speed smoothly transitions from the current value to the target value, avoiding vertical vibration and beam swaying caused by speed change.

[0144] For pressure abrupt change points, a preset exponential smoothing coefficient is invoked. The pressure adjustment direction is determined based on the pressure adjustment amount. If the pressure adjustment amount is positive, it is incremented based on the exponential smoothing coefficient. If the pressure adjustment amount is negative, it is decayed based on the exponential smoothing coefficient, ensuring that the smoothing direction is consistent with the pressure adjustment direction and conforms to the pressure correction target of the load deviation feedback rule. The exponential smoothing coefficient is used to control the increment / decay rate of the pressure adjustment, so that the pressure change conforms to the response characteristics of the SPMT hydraulic suspension. Based on the exponential smoothing formula and the pressure response hysteresis characteristics of the SPMT hydraulic suspension, the pressure change curves under different coefficients are simulated. The coefficient corresponding to no pressure overshoot and a response time of less than 200ms is taken as the optimal value. Finally, the exponential smoothing coefficient is obtained through experimental verification and solidification. In this embodiment, the exponential smoothing coefficient is taken as 0.4.

[0145] For abrupt steering changes, a preset upper limit for steering fine-tuning is invoked. Using this upper limit as a benchmark, the excess adjustment amount is obtained by comparing the steering angle adjustment with the benchmark. And add after the original turning point A series of fine-tuning sampling cycles form a cycle containing 1 minute of the original mutation cycle and The transition segment of the first fine-tuning cycle is determined, and the cumulative fine-tuning ratio to be achieved in each fine-tuning cycle is calculated using the sine term. For example, for the first... The required cumulative fine-tuning percentage is [number] times per fine-tuning cycle. Then, the difference between the cumulative fine-tuning ratio of the current fine-tuning cycle and the cumulative fine-tuning ratio of the previous fine-tuning cycle is used, which is... The single-adjustment ratio for the current fine-tuning cycle is obtained. The single-adjustment amount for the current fine-tuning cycle is then obtained by multiplying the single-adjustment ratio by the over-adjustment amount. This is the difference between the cumulative fine-tuning amount of the current cycle and the cumulative fine-tuning amount of the previous cycle. Therefore, in the last fine-tuning cycle, the overall steering angle adjustment amount meets the target and is consistent with the original target adjustment amount. It was set up by a person skilled in the art based on the actual project;

[0146] For example, if the preset upper limit for steering fine-tuning is 3° and the original steering angle adjustment at the abrupt change point is 5°, then the excess adjustment amount is... In this embodiment, take Then fine-tune the cycle number ;

[0147] The single-step adjustment ratio for the first adjustment cycle is: The single-round adjustment ratio for the second fine-tuning cycle is: ;

[0148] By using the single-adjustment ratio and the over-adjustment amount, the single-adjustment amount for the first fine-tuning cycle was calculated to be 1.41°, and the single-adjustment amount for the second fine-tuning cycle was calculated to be 0.59°.

[0149] This transition phase includes one original mutation cycle and two fine-tuning cycles. The original mutation cycle executes the adjustment to the upper limit of the fine-tuning, i.e. The first fine-tuning cycle is executed. The second fine-tuning cycle is executed. ;

[0150] The smoothed control instructions are then used to replace the segments at the abrupt change points in the original control instruction set, thus forming smoothed control instructions.

[0151] Furthermore, comparing the actual response with the expected response includes:

[0152] The system receives real-time response data uploaded by each SPMT local controller, including actual vehicle speed response, actual pressure response, actual steering response, and auxiliary response. It also uses moving average noise reduction to identify abnormal data. Among these, the auxiliary response is the key auxiliary status feedback data during SPMT transportation, in addition to the main responses of vehicle speed, pressure, and steering. This includes actual steel truss strain response, actual wind load response, and actual SPMT attitude response.

[0153] The expected response is defined as the ideal execution effect of the control command, and the deviation between the actual response and the expected response is calculated; where the ideal execution effect is generated based on the rule base.

[0154] By setting a preset allowable deviation threshold, the deviation value is divided into three levels: slight deviation, moderate deviation, and severe deviation, and a deviation comparison report is generated. The allowable deviation threshold is set through statistical analysis of historical data and includes a first-level deviation threshold and a second-level deviation threshold. For example, the first-level deviation threshold for actual vehicle speed response is 0.2 km / h and the second-level deviation threshold is 0.5 km / h.

[0155] Example 2:

[0156] Please see Figure 4 Another embodiment of the present invention provides a control method for the SPMT (Special Purpose Tunneling and Material Handling) process of steel truss girders, comprising the following steps:

[0157] By utilizing the constructed full-dimensional perception network, the transportation data of the entire process of moving steel truss SPMT is collected in real time and preprocessed, including road elevation data, wind speed and direction data, pressure data, strain data, and vehicle operation data.

[0158] Based on the transport data, road condition disturbance parameters and wind load disturbance parameters are obtained, and the load status of the SPMT transport is determined.

[0159] The system calls a preset rule base, triggers feedforward rules with disturbance parameters to generate pre-action instructions, and triggers feedback rules with load status to generate correction instructions. The system constructs a control instruction set by linearly superimposing these instructions.

[0160] The control command set is smoothed to optimize the command change curve and sent to each SPTM local controller. At the same time, the actual response data after the command is executed is received and compared with the expected response to update the rule base.

[0161] Working principle and effects:

[0162] By building a full-dimensional perception network, the system collects transport data in real time, acquires disturbance parameters, and determines the SPMT load status. Then, it calls a preset rule base and command generator to generate pre-action commands based on disturbance parameters and generates correction commands by comparing the load status with the target status. These commands are then linearly superimposed with weights to form a control command set. This enables early prediction of disturbances and precise correction of operational deviations, generating optimal flexible adjustment commands. This effectively improves transport efficiency, reduces system oscillation risks, and smooths and optimizes the curves of the control command set to avoid operational shocks caused by sudden command changes. The smoothed commands are then sent to the SPMT. Simultaneously, the system receives actual response data and compares it with the expected response, dynamically updates the rule base, and continuously optimizes the control strategy. This enhances the system's adaptability to complex working conditions and ultimately eliminates oscillation problems caused by improper control, ensuring the safety and stability of steel truss transport and achieving real-time flexible and precise control under dynamic disturbances.

[0163] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A control system for the SPMT (Special Purpose Material Handling) transport process of steel truss girders, characterized in that, include: Link acquisition module, judgment module, control module and optimization module; The link acquisition module is used to collect transportation data in real time, including road elevation data, wind speed and direction data, pressure data, strain data, and vehicle operation data; The judgment module obtains road condition disturbance parameters and wind load disturbance parameters through road elevation data and wind speed and direction data. By performing statistical analysis on pressure data and inverse analysis on strain data, it determines the load state of the SPMT being moved. The control module is used to call the preset rule base and instruction generator, generate pre-action instructions based on road condition disturbance parameters and wind load disturbance parameters, generate correction instructions by comparing the load state with the preset target state, perform offline global optimization of the rule base through particle swarm optimization algorithm to obtain instruction superposition weight, and linearly superimpose the pre-action instructions and correction instructions according to the instruction superposition weight to generate a control instruction set. The optimization module is used to smooth the control instruction set to optimize the instruction change curve, and then send the smoothed control instructions to each SPTM. At the same time, it receives the actual response data after the instruction is executed, compares the actual response with the expected response, and updates the rule base.

2. The control system for the SPMT (Special Purpose Turbine Material Handling) transport process of steel truss girders according to claim 1, characterized in that, The steps to obtain road disturbance parameters include: Based on the preset ratio of distance coefficient to real-time vehicle speed, the filter window size is set, and the road elevation data is filtered to obtain smooth elevation data. Set an elevation benchmark for an ideal smooth road surface, and calculate the elevation deviation of each sampling point by combining the elevation measurement value of each sampling point in the smooth elevation data. Obtain the sampling interval of the sensor array and calculate the spatial coordinate position corresponding to each sampling point; By associating the elevation deviation of each sampling point with its corresponding spatial coordinates, a road surface excitation map is constructed. Based on the physical parameters and second-order closed-loop transfer function of the SPMT suspension terminal, the vibration transmission signal of the transfer terminal is calculated. The elevation deviation is subjected to Laplace transform to obtain the elevation frequency domain signal. The vertical acceleration, including the front axle acceleration and the rear axle acceleration, is obtained by performing an inverse Laplace transform on the product of the vibration transmission signal and the elevation frequency domain signal. By calculating the acceleration difference between the front and rear axles, the equivalent mass, and the wheelbase, the pitching moment is calculated. The road excitation map, pitching moment, and vertical acceleration are then integrated to form road condition disturbance parameters.

3. The control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process of steel truss beams according to claim 2, characterized in that, The steps to obtain wind load disturbance parameters include: Smooth the wind speed and direction data and verify the data validity to generate smoothed wind data; Retrieve the wind load factor lookup table to find the benchmark wind load factor corresponding to the current wind speed; Based on the wind direction angle corresponding to the current wind speed, the corresponding wind direction correction factor is selected from the pre-built correction rule table; If no wind direction correction factor is found, select two angles adjacent to the current wind direction angle and their corresponding wind direction correction factors, and obtain the wind direction correction factor for the current wind direction angle through linear interpolation. Based on the wind direction correction factor and the baseline wind load factor, the corrected baseline coefficient is obtained, and then combined with the wind direction angle to be decomposed into the lateral force wind load factor and the lift wind load factor. Obtain the windward area of ​​the steel truss and, in conjunction with the formula for calculating air resistance, calculate the static equivalent wind load, including lateral force wind load and lift wind load. The distance between the point of application of the wind load and the geometric center of the steel truss is obtained. The torsional moment is calculated by combining the static equivalent wind load and the static equivalent wind load, and the wind load disturbance coefficient is formed.

4. The control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process of steel truss beams according to claim 3, characterized in that, Determining the load status of SPMT migration includes: Based on the pressure data, the mean, variance, and range of the pressure at each point are calculated. The eccentric load coefficient is obtained by the ratio of the pressure range to the mean, and the pressure distribution characteristics are obtained. Stress values ​​of key sections of steel truss beams are obtained by processing strain data based on Hooke's law. Based on the geometric parameters of the steel truss section and the bonding design of strain gauges, the stress state of key sections of the steel truss is determined, including tensile stress and compressive stress. Based on vehicle operation data, calculate the attitude offset and attitude change rate during the transfer process; By integrating pressure distribution characteristics, wind load disturbance parameters, and additional forces generated by attitude deviation, the internal force parameters of key sections of the steel truss girder, including axial force, shear force, and bending moment, are calculated based on the static equilibrium equation. Correlation analysis was used to calculate the influence weight of each disturbance parameter on the internal force parameter, and a collaborative correlation map was constructed by associating road condition disturbance parameters, wind load disturbance parameters, and internal force parameters.

5. The control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process of steel truss beams according to claim 4, characterized in that, Determining the load status of SPMT migration also includes: Obtain state determination thresholds based on offline experimental calibration, including off-center load threshold, pressure range threshold, stress threshold, and attitude deviation threshold; Based on the off-center load factor, pressure difference, beam stress value, and attitude offset, the current load state is determined, including load balance state, slight off-center load state, and severe off-center load state. Based on historical data under all transport conditions, instability trend characteristics are extracted, including pressure difference growth rate, stress change rate, and attitude deviation acceleration. Obtain a preset trend threshold and combine it with the characteristics of the instability trend to determine whether there is a potential risk of instability; If any unstable trend feature continues to exceed the trend threshold within 3 consecutive sampling points, it is determined that there is a potential risk of instability, and the risk level is marked accordingly. Based on the assessment of the current load status and potential instability risks, a judgment report is generated for the current SPMT migration.

6. The control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process of steel truss beams according to claim 5, characterized in that: The rule base stores road disturbance feedforward rules, wind load disturbance feedforward rules, load deviation feedback rules, and attitude deviation feedback rules. Each rule has preset trigger conditions, execution actions, and associated parameters, including: The triggering condition for the road disturbance feedforward rule is that the elevation deviation is greater than the local bump threshold and the real-time vehicle speed is greater than the vehicle speed threshold. The execution action is to generate a preset speed curve command to smoothly decelerate to the preset safe vehicle speed. The triggering condition for the wind load disturbance feedforward rule is that the lateral wind load is greater than the wind load threshold and the SPMT is traveling in a straight line. The execution action is to generate a preventive load compensation command to increase the adjustment differential pressure. The load deviation feedback rule is triggered when the pressure range is continuously greater than the pressure range threshold, and the action is to generate a command to enable the pressure PID feedback control algorithm. The attitude deviation feedback rule is triggered when the attitude deviation is continuously greater than the attitude deviation threshold, and the action is to generate a command to enable the attitude PID feedback control algorithm.

7. The control system for the SPMT (Special Purpose Tunneling and Material Handling) transport process of steel truss girders according to claim 6, characterized in that, The steps for generating a control instruction set include: Call the road condition disturbance parameters and wind load disturbance parameters, compare them with the rule base to match the corresponding feedforward rules, and generate pre-action instructions; Based on the safety threshold for feedforward instructions in the rule base, the validity of pre-action instructions is verified to eliminate invalid instructions, and the triggering rule and corresponding disturbance of each instruction are marked. The pressure range and attitude offset are called, and the rule base is compared to match the corresponding feedback rule. The corresponding controller is called to generate correction instructions, including pressure correction instructions and attitude correction instructions. The instruction weights in the rule base are retrieved, and the same type of instructions in the pre-action instructions and correction instructions are linearly superimposed to calculate the adjustment amount, so as to generate the initial control instructions.

8. The control system for the SPMT (Special Purpose Turbine Material Handling) transport process of steel truss beams according to claim 7, characterized in that, The steps for generating a control instruction set include: Obtain an offline working condition dataset containing instruction superposition weights, use the instruction superposition weights as candidate values ​​for particle positions, and construct an objective function based on the transport evaluation index; The working condition parameters in the offline working condition dataset are traversed, the mean of the objective function for all working conditions under different instruction superposition weights is calculated, and the particle position is iteratively updated with the goal of maximizing the mean of the objective function. The optimal instruction superposition weight is obtained and written into the rule base. For the initial control command, commands whose adjustment amount is greater than the command superposition threshold are removed, and the removed commands are replaced with the command superposition threshold as the corresponding adjustment amount. Based on the vehicle speed command and road excitation map in the initial control command, linear interpolation is used to decompose the vehicle speed command into the target vehicle speed value of each road segment and generate the target speed curve. Based on the pressure command in the initial control command and the pressure distribution characteristics of each point, the pressure command is broken down into the target pressure value of each point. Based on the steering angle command, attitude offset, and sampling period in the initial control command, the steering angle command is decomposed into the target steering angle value for each sampling period to form a target steering angle sequence; The target speed curve, target pressure value, and target steering angle sequence are integrated into a control command set, and the execution conditions and execution priority of each command are marked.

9. The control system for the SPMT (Special Purpose Turbine Maneuver) process according to claim 8, characterized in that, The steps to optimize the instruction variation curve include: The system retrieves preset thresholds for vehicle speed smoothing, pressure change, and steering fine-tuning, and identifies abrupt change points in the control commands based on the control command set, including speed abrupt change points, pressure abrupt change points, and steering abrupt change points. For speed change points, instruction segments with one sampling period are set before and after the speed change point to form a transition segment, and a continuous transition speed is generated by linear interpolation. For pressure abrupt change points, the exponential smoothing coefficient is invoked, the direction of pressure adjustment is determined based on the pressure adjustment amount, and the pressure adjustment amount is increased or decreased in combination with the exponential smoothing coefficient. For a sudden steering change point, the excess adjustment amount is obtained by the difference between the steering angle adjustment amount and the upper limit of the steering fine adjustment. A fine adjustment cycle is added after the original sudden steering change point to form a steering transition segment. Using the sine term, the cumulative fine-tuning ratio for each fine-tuning cycle is calculated. The single fine-tuning ratio for the current fine-tuning cycle is obtained by the difference between the cumulative fine-tuning ratios of adjacent fine-tuning cycles. Combined with the overshoot adjustment, the single fine-tuning amount for the current fine-tuning cycle is obtained. The smoothed control instructions are then used to replace the segments at the abrupt change points in the original control instruction set, thus forming smoothed control instructions.

10. A control method for the SPMT (Special Purpose Turbine Mounting) transport process of steel truss girders, implemented based on the control system for the SPMT transport process of steel truss girders as described in any one of claims 1-9, characterized in that, include: By utilizing the constructed full-dimensional perception network, the transportation data of the entire process of steel truss SPMT transportation is collected in real time and preprocessed. Based on the migration data, disturbance parameters are obtained, and the load status of the SPMT migration is determined. The disturbance parameters include road condition disturbance parameters and wind load disturbance parameters. The system calls a preset rule base, triggers feedforward rules with disturbance parameters to generate pre-action instructions, and triggers feedback rules with load status to generate correction instructions. The system constructs a control instruction set by linearly superimposing these instructions. The control command set is smoothed to optimize the command change curve, and then sent to each SPTM local controller for updating based on actual response data.

Images