Method for adaptive control of a curve in a semiconductor production line handling process

Abstract

The present application relates to the technical field of semiconductor manufacturing material handling, and provides a curved track adaptive control method in a semiconductor production line handling process, comprising: obtaining the self geometric parameters of a handling vehicle and the track parameters of a target curved track, and obtaining a relative position state vector of the handling vehicle in the curved track in real time; based on the relative position state vector and the track parameters, and in combination with the real-time load state and inertia feedback of the handling vehicle, performing optimal selection of an inner track or an outer track in the curved track; based on the selected track and the curved track stage where the handling vehicle is located, and in combination with the relative attitude relationship between the front wheel and the rear wheel of the handling vehicle, performing asymmetric speed distribution on the front wheel driving unit and the rear wheel driving unit of the handling vehicle, so as to realize closed-loop correction on the motion track of the handling vehicle. The present application realizes smooth control in the whole cycle of the curved track, effectively suppresses the vehicle body swing, reduces the motor load, and improves the handling positioning accuracy.

Description

Technical Field

[0001] This invention relates to the field of automated material handling technology. Specifically, it relates to a method and system for adaptive control of curves in the material handling process of a semiconductor production line. Background Technology

[0002] Currently, in the semiconductor manufacturing field, the handling accuracy of core materials such as wafers and display panels directly determines the production line yield. In Automated Material Handling Systems (AMHS), overhead rail-mounted transport vehicles (OHT / AGVs) commonly suffer from speed control mismatch when operating on curves. Traditional solutions typically employ control logic that forces the front and rear wheels to maintain the same speed, or simply reduce speed based on a fixed curvature. Because the wheelbase of the semiconductor transport vehicle is fixed, and the track radii differ between the inner and outer sides of the curve, the actual arc lengths of the front and rear wheels' trajectories are inconsistent during the transition phase between entering and exiting the curve.

[0003] This kinematic asynchrony generates additional lateral shear forces, causing the vehicle to sway. For nanoscale semiconductor processes, this swaying can easily lead to micro-vibrations inside the wafer pod (FOUP), causing microcracks or chip detachment. Furthermore, existing control methods lack fine-grained segmentation of the entire curve's cycle, making it impossible to correct the operating state in real time using high-precision relative position vectors. This results in excessively high servo motor load and wheel slippage when the transport vehicle is operating on curves, severely impacting equipment lifespan and handling safety. Summary of the Invention

[0004] To address the problems of vehicle swaying, excessive motor load, and insufficient positioning accuracy caused by mismatch in front and rear wheel speed control when the transport vehicle is running on curves in the existing technology, this application proposes a curve adaptive control method, system, and medium for semiconductor production line transport processes. By combining the transport vehicle's dynamic model and relative state vectors for stage division and correction, smooth control of the entire curve cycle can be achieved.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: An adaptive control method for curves in the material handling process of a semiconductor production line includes: acquiring the geometric parameters of the material handling vehicle and the track parameters of the target curve, and acquiring the relative position state vector of the material handling vehicle in the curve in real time. Based on the relative position state vector and track parameters, combined with the real-time load state and inertial feedback of the material handling vehicle, an optimal selection is made between the inner and outer tracks of the curve. Based on the selected track and the stage of the curve in which the material handling vehicle is located, and combined with the relative attitude relationship between the front and rear wheels of the material handling vehicle, asymmetrical speed distribution is performed on the front-wheel drive unit and the rear-wheel drive unit of the material handling vehicle to achieve closed-loop correction of the material handling vehicle's motion trajectory.

[0006] The above scheme achieves refined division of the curve operation phase and optimal selection of the track by introducing relative position state vector and real-time load state, and eliminates the kinematic interference of the front and rear wheels by using asymmetric speed distribution, thereby effectively suppressing vehicle body sway, reducing motor load, and improving handling and positioning accuracy.

[0007] As an optional implementation, the steps for optimally selecting the inner or outer track of a curve, based on the real-time load status and inertial feedback of the transport vehicle, include: obtaining the current load weight W of the transport vehicle and the real-time vibration amplitude A of the chassis; setting a load safety threshold Wth and a vibration safety threshold Ath; when W > Wth and A ≤ Ath, prioritizing the outer track of the curve to utilize the larger turning radius and reduce the risk of rollover; when W ≤ Wth and A > Ath, prioritizing the inner track with lower vibration transmission to suppress micro-vibrations.

[0008] The above solution achieves adaptive track selection by quantifying load weight and vibration amplitude. Under heavy load, it prioritizes anti-rollover safety, and under light load and vibration, it prioritizes suppressing micro-vibrations, thereby improving the system's adaptability.

[0009] As an optional implementation, the optimal selection of the inner or outer track of the curve, based on the real-time load status and inertial feedback of the transport vehicle, further includes the step of: calculating the comprehensive risk index R=α⋅W / W when W>Wth and A>Ath. th +β⋅A / A th α and β are preset weighting coefficients. If R exceeds the preset risk limit, emergency braking is triggered and track selection is suspended until W or A falls back to a safe range.

[0010] The above solution introduces a comprehensive risk index assessment mechanism, which actively triggers emergency braking under extreme working conditions of heavy load and high vibration, avoiding potential safety accidents and further ensuring the reliability of the handling process.

[0011] As an optional implementation, the steps of obtaining the transport vehicle's own geometric parameters and the track parameters of the target curve include, and in real time, obtaining the relative position state vector of the transport vehicle in the curve, including: determining the wheelbase between the front and rear wheels of the transport vehicle, the inner track curvature of the curve, the center track curvature of the curve, and the preset running speed of the transport vehicle. By real-time monitoring of the displacement changes of the front wheel drive unit, combined with the curve's geometric constraints, the relative state vector characterizing the real-time position of the transport vehicle in the curve is calculated.

[0012] The above scheme provides a precise data foundation for subsequent refined control by calculating the relative state vector through displacement monitoring and geometric constraints, thus ensuring the accuracy of the control logic.

[0013] As an optional implementation, the steps for selecting the inner or outer track of a curve based on the relative position state vector and track parameters, combined with the real-time load state and inertial feedback of the transport vehicle, include: determining whether the transport vehicle is in one of the three stages of curve entry transition, curve section, or curve exit transition based on the relative position state vector and track parameters, including: when the relative state vector indicates that the transport vehicle has just entered the curve and there is a difference in track type between the front and rear wheels, it is determined that the transport vehicle is in the curve entry transition section.

[0014] When the relative state vector indicates that the transport vehicle is completely within the curve area and the track type of the front and rear wheels is consistent, the transport vehicle is determined to be in the curve section. When the relative state vector indicates that the transport vehicle is about to leave the curve and the track type of the front and rear wheels changes again, the transport vehicle is determined to be in the curve exit transition section.

[0015] The above scheme accurately determines the curve stage of the transport vehicle by using relative state vectors, providing a prerequisite for implementing differentiated control strategies at different stages and ensuring the continuity and smoothness of the control process.

[0016] As an optional implementation, when the transport vehicle is in the transition section before entering a curve, an asymmetric speed distribution is performed on the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle. This includes: calculating a first geometric angle between the front wheels and the center of the currently selected curve based on the relative attitude relationship; calculating a second geometric angle formed by the line connecting the front and rear wheels based on the wheelbase and current displacement of the transport vehicle; determining a straight-segment travel compensation factor for the rear wheels based on the difference between the first and second geometric angles; and reducing the output speed of the rear-wheel drive unit according to the compensation factor, while maintaining the speed of the front-wheel drive unit matching the curvature of the currently selected curve.

[0017] The above solution compensates for the speed of the rear wheels during the cornering transition, eliminating the lateral shear force generated when the front wheels enter the corner while the rear wheels are still on the straight section, effectively suppressing vehicle sway during cornering.

[0018] As an optional implementation, when the transport vehicle is on a curve, asymmetrical speed distribution is performed on the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle. This includes: calculating the theoretical linear velocities of the front and rear wheels in circular motion based on the ratio of the curvature radii of the inner track and the center track of the currently selected curve. The front-wheel drive unit and the rear-wheel drive unit are then controlled to output speed values ​​corresponding to the curvature of their respective tracks, so that the transport vehicle's center of gravity trajectory aligns with the curve's center.

[0019] The above solution uses differential control based on the radius of curvature in curves to ensure vehicle stability and avoid tire slippage and motor overload caused by forced uniform speed.

[0020] As an optional implementation, when the transport vehicle is in the transition section after exiting a curve, asymmetrical speed distribution is performed on the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle. This includes: calculating the direction correction angle for the front wheels transitioning from curve to straight-line travel based on the geometric relationship between the current position of the transport vehicle and the curve exit; calculating the angle between the rear wheels and the line connecting the center of the currently selected curve, based on the geometric constraints that the rear wheels are still within the curve; determining the interpolation coefficient for the front wheels' transition from tangential velocity to linear velocity using the coupling relationship between the direction correction angle and the line angle; and smoothly transitioning the output speed of the front-wheel drive unit according to the interpolation coefficient, while maintaining the speed of the rear-wheel drive unit matching the curvature of the currently selected curve.

[0021] The above solution provides a smooth transition correction for the front wheels during the corner exit phase, avoiding the impact of sudden acceleration or deceleration and ensuring the stability of the corner exit process.

[0022] Furthermore, this invention also provides a curve adaptive control system for the handling process in a semiconductor production line, comprising: a track selection module configured to decide whether to select the inner or outer track of the curve based on the real-time load status and inertial feedback of the transport vehicle; and a parameter acquisition module configured to acquire the transport vehicle's own geometric parameters, the track parameters of the target curve, and the real-time relative position vector.

[0023] The stage determination module is configured to determine whether the transport vehicle is in the entry-to-curve transition section, the curve section, or the exit-to-curve transition section based on the relative position state vector and track parameters. The execution control module is configured to perform asymmetric speed distribution and closed-loop correction on the front-wheel drive unit and the rear-wheel drive unit based on the selected track and the determined stage.

[0024] The above system achieves full-process adaptive control of the transport vehicle's curved operation through the coordinated work of its various modules, effectively solving the vibration and wear problems caused by control mismatch in existing technologies.

[0025] In addition, the present invention also provides a computer-readable storage medium in which a computer program, when executed by a processor, implements a curve adaptive control method for the handling process of a semiconductor production line as described in any of the above solutions.

[0026] The beneficial effects of this invention are as follows: The adaptive control method, system, and medium for curved sections in the handling process of a semiconductor production line provided by this invention have the following main advantages: First, by acquiring the relative position vector of the transport vehicle in the curve in real time, and combining the load status and inertial feedback to select the optimal track, the system achieves accurate perception and adaptive decision-making of the curve operation environment, avoiding the limitations of traditional single control logic.

[0027] Secondly, the curve process is divided into the entry transition section, the curve section, and the exit transition section. Asymmetrical speed distribution is carried out according to the geometric constraint relationship between the front and rear wheels at different stages. From the perspective of kinematics, the lateral shear force caused by the inconsistent arc length of the front and rear wheel trajectories is eliminated, and the body sway amplitude is limited to within the micro-vibration threshold, effectively preventing wafer microcracks or particle shedding.

[0028] Finally, by using dynamic speed distribution based on geometric relative position vectors, the "internal friction" generated in traditional forced uniform speed control is avoided, which significantly reduces the peak torque requirement of the servo motor and the risk of wheel slippage, extends the service life of the equipment, and ensures the precise stopping of the transport vehicle after high-speed cornering.

[0029] The following section will further explain the characteristics, technical features, advantages, and implementation of the adaptive control method and system for curves in the semiconductor production line handling process in a clear and easy-to-understand manner, with the aid of accompanying figures. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating an adaptive control method for curves in a semiconductor production line handling process, as described in one embodiment of the present invention.

[0031] Figure 2 This is a schematic diagram illustrating the composition of a curve adaptive control system based on the handling process of a semiconductor production line in one embodiment of the present invention.

[0032] Figure 3 This is a schematic diagram illustrating the shape of a curve in one embodiment of the present invention. Detailed Implementation

[0033] To provide a clearer understanding of the technical features, objectives, and effects of the invention, specific embodiments of the invention are now described with reference to the accompanying drawings. In the drawings, the same reference numerals indicate components with the same or similar structures but the same function.

[0034] In this document, "illustrative" means "serving as an example, illustration, or description," and any illustrations or embodiments described herein as "illustrative" should not be construed as a more preferred or advantageous technical solution. For the sake of brevity, each figure only schematically shows the parts relevant to this exemplary embodiment, and they do not represent the actual structure or true proportions of the product.

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0037] In implementation method one, such as Figure 1 As shown, Figure 1 This is a flowchart illustrating an adaptive control method for curves in a semiconductor production line transport process, as described in one embodiment of the present invention. The method aims to solve the problems of vehicle swaying, excessive motor load, and insufficient positioning accuracy caused by mismatched front and rear wheel speed control when the transport vehicle is operating on curves in the prior art. By introducing a relative position state vector and an asymmetric speed distribution strategy, smooth control throughout the entire curve cycle is achieved.

[0038] Specifically, the method includes the following steps: Step S100: Obtain the transport vehicle's own geometric parameters and the track parameters of the target curve, and obtain the relative position state vector of the transport vehicle in the curve in real time.

[0039] like Figure 3 As shown, the vehicle's own geometric parameters include fixed dimensional data such as the wheelbase and track width of the front wheels 21 and rear wheels 22 of the transport vehicle 20. The track parameters of the target curve include geometric feature data such as the inner track curvature 11, outer track curvature 13, center track curvature 12, and curve arc length of the curve 10. The "relative position state vector" referred to in this embodiment is a comprehensive variable used to characterize the real-time position and attitude of the transport vehicle in the curve coordinate system. This vector not only includes the coordinate position of the transport vehicle's center but also implicitly contains attitude information such as the vehicle's deflection angle relative to the track tangent and the relative positional relationship between the front and rear wheels.

[0040] By monitoring the displacement changes of the front-wheel drive unit or the rear-wheel drive unit in real time, and combining this with the geometric constraints of the curve, the state vector can be dynamically calculated. This vectorized description method enables the system to handle the control problems of transport vehicles with different radii of curvature and wheelbase specifications within a unified mathematical framework, providing a data foundation for subsequent refined control.

[0041] Step S200: Based on the relative position state vector and track parameters, combined with the real-time load status and inertial feedback of the transport vehicle, the optimal selection of the inner or outer track of the curve is made.

[0042] In semiconductor production lines, the load conditions (such as wafer box weight) and operating conditions (such as vibration amplitude) of transport vehicles are dynamically changing. If the track is selected based solely on static map information, it may lead to an increased risk of overturning of heavily loaded vehicles at small-radius curves, or the transmission of micro-vibrations when lightly loaded, high-vibration vehicles pass through seams.

[0043] This step introduces real-time load status and inertial feedback as decision variables to construct a dynamic track selection mechanism. Specifically, when the transport vehicle is under heavy load, the system tends to choose the outer track with a larger radius to reduce centrifugal force. When the transport vehicle is under vibration-sensitive conditions, the system tends to choose the inner track to shorten the curve path. This optimal selection is not a simple static programming, but a dynamic decision based on real-time state vectors, ensuring that the transport vehicle is in optimal kinematic initial conditions before entering the curve.

[0044] Step S300: Based on the selected track and the curve stage where the transport vehicle is located, and combined with the relative attitude relationship between the front and rear wheels of the transport vehicle, asymmetrical speed distribution is performed on the front wheel drive unit and the rear wheel drive unit of the transport vehicle to achieve closed-loop correction of the transport vehicle's motion trajectory.

[0045] This is the core step of this implementation method. Traditional control methods often force the front and rear wheels to travel at the same speed, ignoring the difference in the actual arc length swept by the front and rear wheels during cornering of a rigid vehicle. The "asymmetric speed distribution" proposed in this implementation method refers to assigning different target speeds to the front-wheel drive unit and the rear-wheel drive unit within the same control cycle. This differentiated speed setting is calculated based on the relative attitude relationship between the front and rear wheels. For example, during the cornering phase, the front wheels have entered the cornering trajectory while the rear wheels are still on a straight trajectory. If forced to travel at the same speed, the rear wheels will be "dragged" by the front wheels, generating lateral shear force.

[0046] By asymmetrically distributing the speed, the rear wheel speed is actively reduced or the front wheel speed is adjusted, ensuring that the linear velocities of the front and rear wheels match their respective trajectory curvatures, thus eliminating interference forces in terms of kinematics. This process is closed-loop, meaning the system continuously adjusts the speed distribution strategy based on the latest relative position vector until the transport vehicle smoothly exits the curve. It should be understood that the specific algorithm for asymmetric speed distribution will adaptively adjust according to different curve stages (entry transition section, curve section, exit transition section), but the core logic remains the same: maintaining the smoothness of the transport vehicle's center of gravity trajectory and the stability of its posture.

[0047] Through the above steps, this implementation method establishes a complete adaptive control logic for curves. From the comprehensiveness of data acquisition (relative position state vector), to the dynamic adaptability at the decision level (track optimization), and then to the kinematic decoupling at the execution level (asymmetric speed distribution), a closed-loop control system from perception to execution is formed, effectively solving the problems of micro-vibration suppression and motor load optimization in semiconductor handling.

[0048] In Implementation Method Two, based on Implementation Method One, this implementation method elaborates on the specific logic of track optimization selection. This logic aims to solve the problem of insufficient operational stability caused by prior art that only plans paths based on static maps and ignores the real-time dynamic status of the transport vehicle.

[0049] Specifically, step S200, which combines the real-time load status and inertial feedback of the transport vehicle to select the optimal inner or outer track for the curve, includes the following sub-steps: Step S201: Obtain the current load weight W of the transport vehicle and the real-time vibration amplitude A of the chassis. The load weight W can be obtained using a weighing sensor installed at the bottom of the vehicle body, and the vibration amplitude A can be obtained using an accelerometer or gyroscope installed on the chassis. It should be understood that the vibration amplitude A can be the peak value of vertical acceleration or the horizontal sway amplitude, depending on the sensor installation method and the calibration strategy of the control system.

[0050] Step S202, set the load safety threshold W th Vibration safety threshold A th These two thresholds are system-preset safety boundary values ​​used to define whether the transport vehicle is currently in a safe operating condition. For example, in one application scenario, the load safety threshold W... th The vibration safety threshold A can be set to 80% of the rated load of the pallet truck. th It can be set to the permissible micro-vibration limit value for semiconductor wafer handling (e.g., 0.3G).

[0051] Step S203: Based on the comparison results of the load weight W and vibration amplitude A with their respective thresholds, a differentiated track selection strategy is executed. This implementation constructs a defense depth for "normal operating conditions" and "abnormal operating conditions," and the specific logic is as follows: When W>W th And A≤A thWhen the transport vehicle is determined to be in a stable, heavy-load condition, the system prioritizes selecting the outer track of the curve. This is because under heavy load conditions, the transport vehicle's center of gravity is higher or its inertia is greater, resulting in a larger centrifugal force when cornering. According to the centrifugal force formula F=mv² / r, with a constant speed v, selecting the outer track with a larger radius r can significantly reduce the centrifugal force, thereby effectively reducing the risk of tipping over. This selection strategy reflects the principle of "safety first," using geometric path optimization to compensate for dynamic risks.

[0052] When W≤W th And A>A th When the transport vehicle is determined to be under light load and high vibration conditions, the system prioritizes the inner track, which transmits less vibration. In semiconductor production lines, minor unevenness in the track joints or foundation can cause high-frequency vibrations in the vehicle body. The inner track typically has a smaller radius of curvature, and although the centrifugal force is slightly larger, it is manageable under light load conditions. More importantly, in some layouts, the inner track may avoid the joints of the outer track, or the supporting structure of the inner track may have higher rigidity, thus effectively suppressing the transmission of micro-vibrations and preventing microcracks in the wafer due to resonance.

[0053] Furthermore, this implementation also introduces an emergency response mechanism for extreme operating conditions. When W>W th And A>A th When the transport vehicle is simultaneously under heavy load and high vibration, it usually means that the equipment is operating under extremely harsh conditions or that there is a potential risk of mechanical failure. In this situation, simple track selection is insufficient to eliminate safety hazards. Therefore, this implementation calculates the comprehensive risk index R = α⋅W / W. th +β⋅A / Ath, where α and β are preset weighting coefficients.

[0054] Specifically, the values ​​of α and β typically range from 0.4 to 0.6, and the specific values ​​can be calibrated according to the actual safety level of the production line. For example, if the focus is on preventing rollover, α can be set to 0.6 and β to 0.4. If the focus is on preventing vibration damage, α can be set to 0.4 and β to 0.6. The comprehensive risk index R quantifies the risk of the two dimensions into a single value through normalization. If R exceeds the preset risk limit value (e.g., R_lim=1.5), emergency braking is triggered and track selection is suspended until W or A falls back to the safe range.

[0055] The core logic of this design lies in a "defensive safety strategy." Under extreme conditions of heavy load and high vibration, the dynamic characteristics of the transport vehicle approach their physical limits. Forcibly switching tracks or maintaining operation could lead to rollover or equipment damage. In this situation, the system proactively sacrifices operational efficiency and chooses emergency braking to control the safety risk on the spot, preventing the accident from escalating. This provides a last line of defense for the system's reliability, preventing loss of control due to abnormal sensor data or sudden road conditions.

[0056] In Implementation Method 3, based on Implementation Method 1 or 2, this implementation method elaborates on the specific structure of parameter acquisition and the calculation process of the relative position state vector, and performs refined judgment on the curve operation stage based on this. This process is the key step in transforming abstract control logic into specific execution instructions.

[0057] Specifically, step S100, which involves acquiring the transport vehicle's own geometric parameters and the track parameters of the target curve, and obtaining the transport vehicle's relative position vector within the curve in real time, includes the following steps: Step S101 determines the wheelbase of the front and rear wheels of the transport vehicle, the curvature of the inner track of the curve, the curvature of the center track of the curve, and the preset operating speed of the transport vehicle. These parameters are the basic static data for the control system to perform geometric calculations. Among them, the wheelbase is an inherent property of the transport vehicle, which determines the degree of geometric interference between the front and rear wheels when cornering. The track curvature defines the geometry of the curve and directly affects the calculation of the theoretical linear velocity. It should be understood that the preset operating speed can be the rated speed set by the system or a target speed dynamically adjusted according to production line scheduling instructions.

[0058] Step S102 involves real-time monitoring of the displacement changes of the front-wheel drive unit, combined with the geometric constraints of the curve, to calculate the relative state vector representing the real-time position of the transport vehicle in the curve. The "relative state vector" referred to in this embodiment physically represents the real-time position and attitude of the transport vehicle relative to the geometric center of the curve. Specifically, the front-wheel drive unit is typically equipped with a high-precision encoder, capable of providing real-time feedback on the rotation angle or linear displacement of the front wheels.

[0059] The control system reads the displacement data and, combined with known curve geometry constraints (such as the coordinates of the curve's starting point and radius of curvature), calculates the real-time coordinates of the transport vehicle's center in the curve coordinate system, the vehicle's yaw angle, and the angles between the front and rear wheels relative to the curve's tangent using a forward kinematics algorithm. These parameters together constitute the relative state vector. This encoder-based displacement vector calculation method eliminates the need for expensive external positioning systems (such as LiDAR SLAM), achieving high-precision relative positioning solely through feedback from the transport vehicle's own drive unit, thus reducing system costs and improving reliability.

[0060] Furthermore, based on the aforementioned relative position state vector and track parameters, this embodiment provides a specific method for determining the curve stage, dividing the curve operation process into three consecutive stages with distinct physical characteristics. The step of selecting the inner or outer track in step S200, or subsequent control steps, all rely on accurate stage determination. The specific determination logic is as follows: When the relative state vector represents the transport vehicle just entering a curve and there is a difference in track type between the front and rear wheels, the transport vehicle is determined to be in the curve transition phase. Here, "track type difference" means that the front wheels are already traveling on a curved track with curvature, while the rear wheels are still traveling on a straight track. In this phase, the front wheels are constrained by the curve's geometry and undergo circular motion, while the rear wheels maintain straight-line motion. The kinematic models of the two are fundamentally different, and without differentiated control, enormous lateral shear forces will be generated.

[0061] When the relative state vector indicates that the transport vehicle is completely within the curve area and the front and rear wheel track types are consistent, the transport vehicle is determined to be in a curve segment. At this time, both the front and rear wheels are on the curve track, and the kinematic model is unified as circular motion. However, the difference in the inner and outer track radii causes the front and rear wheels to travel at different arc lengths, requiring steady-state differential speed control.

[0062] When the relative state vector indicates that the transport vehicle is about to leave the curve and there is a difference in track type between the front and rear wheels, it is determined that the transport vehicle is in the curve exit transition section. At this time, the front wheels have left the curve and entered the straight track, while the rear wheels are still in the curve. The kinematic characteristics of this stage are symmetrical to but opposite to those of the curve entry transition section. The front wheels need to smoothly transition from circular motion to straight motion. If not properly controlled, it is very easy to produce a "fishtailing" phenomenon.

[0063] The reason this implementation divides the curve into the three stages mentioned above, rather than a simple "curve / straight" dichotomy, is fundamentally because the kinematic constraints of the front and rear wheels are completely different in different stages. In the transition sections before and after the curve, the front and rear wheels are subject to linear constraints and circumferential constraints, respectively. This mixed constraint state is the root cause of kinematic interference.

[0064] Only by accurately identifying these transition stages through relative state vectors can we employ targeted asymmetric velocity allocation strategies (as in Implementation Method 1) to eliminate interference forces at the physical level. This stage division is a prerequisite for achieving full-cycle smooth control and provides a logical basis for using different control algorithms for different stages in subsequent implementation methods.

[0065] In Implementation Method Four, based on Implementation Method Three, this implementation method elaborates on the specific control strategy when the transport vehicle is in the transition section before entering a curve. This stage is the stage with the most intense kinematic interference during curve operation and is also the key source of micro-vibrations.

[0066] Specifically, when the transport vehicle is in the transition section before entering a curve, the steps for asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle include: Step S301: Based on the relative attitude relationship, calculate the first geometric angle between the front wheel and the currently selected center of the curve. During the transition phase into the curve, the front wheel has entered the curve trajectory, and its position can be determined by the encoder displacement data of the front wheel drive unit. Assuming the center radius of the curve is R, the inner track radius is r, and the displacement of the front wheel relative to the starting point of the curve is S, the arc length of the front wheel moving along the inner track can be calculated. The first geometric angle is the central angle corresponding to this arc length, representing the absolute angular position of the front wheel in the curve coordinate system. This angle is the basis for calculating the theoretical linear velocity of the front wheel, ensuring that the front wheel can travel at a tangential velocity matching the curvature.

[0067] Step S302: Based on the wheelbase and current displacement of the transport vehicle, calculate the second geometric angle formed by the line connecting the front and rear wheels. Since the rear wheels are still on a straight segment, there is an angle between the line connecting the centers of the front and rear wheels and the tangent direction of the curve. This second geometric angle reflects the degree of deflection of the vehicle's attitude relative to the track. In the specific calculation process, this angle can be calculated using trigonometric functions based on the position of the central angle of the front wheels, the wheelbase L, and the geometric projection relationship. This angle intuitively reflects the trend of the rear wheels being "dragged" away from the straight trajectory by the front wheels.

[0068] Step S303: Determine the straight-line travel compensation factor for the rear wheels based on the difference between the first and second geometric angles. This embodiment introduces the concept of a "straight-line travel compensation factor," which physically quantifies the lateral force component generated when the rear wheels are "dragged" by the front wheels on a straight segment. In traditional forced uniform speed control, the front wheels travel in an arc while the rear wheels travel in a straight line; their displacement increments are geometrically unequal, and forcing uniform speed can lead to lateral slippage of the rear wheels. This embodiment constructs a compensation coefficient by calculating the difference between the first and second geometric angles. This coefficient maps the circular motion increment of the front wheels to the linear motion increment of the rear wheels, thereby mathematically eliminating kinematic interference.

[0069] Step S304 involves adjusting the output speed of the rear-wheel drive unit based on a compensation factor, while simultaneously maintaining the speed of the front-wheel drive unit in line with the curvature of the selected curve. Specifically, the target speed of the rear wheels is corrected to the base speed multiplied by the compensation factor (typically a value less than 1), achieving "asymmetric" speed reduction. The front wheels, based on the curve's curvature radius, output the corresponding linear speed according to the laws of circular motion. Through this differential control, the rear wheels no longer passively follow the arc length increment of the front wheels, but instead travel according to their own required displacement increment on straight sections, fundamentally eliminating lateral shear forces and ensuring stable vehicle posture during cornering.

[0070] It should be understood that the calculation process for the first and second geometric angles is derived based on rigid body kinematic constraints. The specific formulas can be adaptively adjusted according to the specific geometric parameters of the curve (such as radius of curvature and wheelbase), and this embodiment does not impose any limitations on this. Through this refined geometric calculation and compensation control, this embodiment effectively solves the vibration problem caused by the difference in front and rear wheel trajectories during the curve transition section, providing technical assurance for the smooth handling of semiconductor wafers.

[0071] In Implementation Method 5, based on Implementation Method 3 or 4, this implementation method elaborates on the specific control strategy when the transport vehicle is in a curve. This stage is the longest steady-state operation stage during the transport vehicle's curve, and its control objective is to maintain the stability of the vehicle's posture and eliminate kinematic interference caused by the difference in the radii of the inner and outer wheel trajectories.

[0072] Specifically, when the transport vehicle is on a curve, the steps for asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle include: Step S401: Based on the ratio of the curvature radii of the inner track and the center track of the currently selected curve, calculate the theoretical linear velocities of the front and rear wheels in circular motion. In the curve section, both the front and rear wheels of the transport vehicle have fully entered the curve area, and the transport vehicle as a whole is in circular motion. According to the principles of rigid body kinematics, when different points on the same rigid body move in a circle around the same center, their linear velocity is directly proportional to the distance (i.e., radius) from that point to the center.

[0073] If the front and rear wheels are forced to travel at the same speed, the difference in the inner and outer track radii (e.g., the inner track radius r is smaller than the outer track radius R) will cause the inner wheel to slip due to insufficient linear velocity, or the outer wheel to be dragged due to excessive linear velocity. This implementation eliminates this kinematic mismatch from a physical perspective by introducing a proportional relationship between the radii of curvature to calculate the theoretical linear velocities of the front and rear wheels.

[0074] Step S402: Control the front-wheel drive unit and the rear-wheel drive unit to output speed values ​​corresponding to the curvature of their respective tracks, so that the center of gravity trajectory of the transport vehicle matches the center of the curve. Specifically, assume the rated center speed of the transport vehicle is V0, the radius of the curve center is R, and the radius of the inner track is r. If the transport vehicle chooses to travel on the inner track, the target speed of the front wheel (assuming it is the inner wheel) should be adjusted to V_inner = V0 * (r / R), and the target speed of the rear wheel (assuming it is the outer wheel) should be adjusted to V_outer = V0 * ((r+L) / R), where L is a geometric parameter related to the wheelbase. Through this precise speed distribution, both the front and rear wheels of the transport vehicle can run at a tangential speed that matches their track radius, avoiding lateral sliding friction between the tires and the track.

[0075] It should be understood that "asymmetry" here refers not only to the possible differences in the speed values ​​of the front and rear wheels, but also to the fact that their control logic is independently calculated based on their respective geometric constraints. The core effect of this control strategy is that it ensures that the center of gravity trajectory of the transport vehicle strictly follows the centerline of the curve (i.e., "following the curve's center"), avoiding centrifugal trajectory deviation or centripetal trajectory contraction caused by speed mismatch. This ensures the stability of the wafer cassette during steady-state cornering and further reduces the risk of micro-vibrations.

[0076] This implementation method solves the problems of tire wear and motor overload caused by traditional forced uniform speed control by implementing differential control based on the radius of curvature in the curve section, and realizes the steady and smooth operation of the transport vehicle in the curve.

[0077] In Implementation Method Six, based on Implementation Methods Three to Five, this implementation method elaborates on the specific control strategy when the transport vehicle is in the transition section after exiting a curve. This stage is the final critical link for the transport vehicle to leave the curve and enter a straight line. If not controlled properly, it is very easy to cause a "tail-wagging" phenomenon or a sudden speed change impact.

[0078] Specifically, when the transport vehicle is in the transition section after exiting a curve, the steps for asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle include: Step S501: Based on the geometric relationship between the current position of the transport vehicle and the exit of the curve, calculate the direction correction angle for the front wheels as they transition from the curve to straight-line travel. At the exit transition section, the front wheels have already passed the curve tangent point and entered the straight track, while the rear wheels remain within the curve. At this point, the velocity vector direction of the front wheels needs to smoothly transition from the curve tangent direction to the straight track direction.

[0079] The direction correction angle is the angle between these two directions, and its physical significance lies in quantifying the amount of attitude adjustment required for the front wheels. In the specific calculation process, the system calculates this correction angle through geometric projection based on the displacement of the front wheels relative to the exit of the curve and the extension direction of the straight track.

[0080] Step S502: Based on the geometric constraints that the rear wheels are still within the curve, calculate the angle between the line connecting the rear wheels and the currently selected center of the curve. At this time, the rear wheels are still in circular motion, and their position can be determined by the displacement feedback of the rear wheel drive unit combined with the curve radius. The angle represents the remaining travel angle of the rear wheels within the curve, and this angle directly determines the theoretical linear velocity required by the rear wheels.

[0081] Step S503: Utilizing the coupling relationship between the direction correction angle and the angle between the connecting lines, the interpolation coefficient for the front wheel's transition from tangential velocity to linear velocity is determined. In this embodiment, the "coupling relationship" refers to the fact that, geometrically, the linearization process of the front wheel and the remaining travel of the rear wheel in the curve occur synchronously, with a strict functional correspondence between the two in time and space. The interpolation coefficient is a variable between 0 and 1, used for linear or nonlinear interpolation between the tangential velocity and linear velocity of the front wheel. For example, when the rear wheel just enters the exit phase of the curve, the interpolation coefficient is close to 0, and the front wheel mainly maintains tangential velocity. As the rear wheel gradually leaves the curve, the interpolation coefficient gradually increases to 1, and the front wheel speed completely transitions to linear velocity.

[0082] Step S504 involves smoothly transitioning the output speed of the front-wheel drive unit based on the interpolation coefficients, while maintaining the speed of the rear-wheel drive unit in line with the curvature of the currently selected curve. This control logic contrasts sharply with the entry-to-curve transition strategy in Implementation Method 4: the entry-to-curve transition primarily corrects the rear-wheel speed to eliminate the "drag" effect, while the exit-to-curve transition primarily corrects the front-wheel speed to eliminate the risk of "fishtailing".

[0083] By smoothly adjusting the interpolation coefficients, the front wheel speed is no longer abruptly abrupt, but rather transitions continuously with changes in geometric position, thus ensuring a smooth recovery of the vehicle's attitude. The rear wheels continue to follow the control logic of the cornering section, outputting a speed that matches the curvature until the vehicle has completely exited the corner.

[0084] It should be understood that the interpolation coefficients in this embodiment can be calculated using various methods, such as linear interpolation, polynomial interpolation, or sinusoidal interpolation based on kinematic models, depending on the required level of stability. Through this asymmetric velocity distribution based on geometric coupling, this embodiment effectively solves the dynamic impact problem caused by the switching of front and rear wheel motion constraints during the corner exit phase, thus perfecting the smooth control closed loop of the transport vehicle throughout its entire lifecycle.

[0085] like Figure 2 As shown in Embodiment Seven, this embodiment provides a curve adaptive control system for the handling process in a semiconductor production line. This system is used to implement the control methods of the above embodiments. The system adopts a modular hardware architecture design, mapping complex control logic to specific physical execution units, and realizing closed-loop control of "perception-decision-execution".

[0086] Specifically, the system includes a track selection module 101, a parameter acquisition module 201, a stage determination module 301, and an execution control module 401. These four modules can be software functional units integrated into the same controller (such as a PLC or industrial computer) or distributed independent hardware boards. This embodiment does not impose any restrictions on this.

[0087] The parameter acquisition module is the system's sensing front end, configured to acquire the transport vehicle's own geometric parameters, the target curve's track parameters, and the real-time relative position vector. At the hardware level, the parameter acquisition module mainly includes various sensor units installed on the transport vehicle. For example, its own geometric parameters (such as wheelbase) can be pre-stored in non-volatile memory. The target curve's track parameters can be obtained from the production line scheduling system via a communication interface.

[0088] The real-time relative position state vector is calculated by monitoring the encoder feedback signals of the front-wheel drive unit and the rear-wheel drive unit in real time, combined with the solution of cornering geometric constraints. This module converts analog signals or discrete position pulses from the physical world into digital state vector data, providing a data foundation for subsequent decision-making.

[0089] The track selection module is the core of the system's decision-making. It is configured to select either the inner or outer track of the curve based on the real-time load status and inertial feedback of the transport vehicle. The hardware of this module is typically a high-performance microprocessor or digital signal processor (DSP). During operation, the track selection module receives the load weight W from the weighing sensor and the vibration amplitude A from the acceleration sensor in real time, and executes the threshold judgment logic and risk index calculation described in Implementation Method Two. The decision signal output by this module is a track selection command (inner or outer), which determines the transport vehicle's path planning before entering the curve, mitigating the risks of overloaded rollover or lightly loaded vibration at a macroscopic level.

[0090] The stage determination module is the logical hub of the system. It is configured to determine whether the transport vehicle is in the entry-to-turn transition section, the curve section, or the exit-to-turn transition section based on the relative position state vector and track parameters. This module mainly runs a logical judgment algorithm. Its input is the relative position state vector provided by the parameter acquisition module, and its output is the current curve stage identifier. By comparing the track type differences (straight line or curve) of the front and rear wheels in real time, this module can accurately identify the spatiotemporal position of the transport vehicle in the curve coordinate system, thereby triggering control strategies for different stages. This process is similar to the jump logic of a state machine, ensuring accurate timing of control strategy switching.

[0091] The execution control module is the system's execution terminal. It is configured to perform asymmetric speed distribution and closed-loop correction on the front-wheel drive unit and rear-wheel drive unit based on the selected track and decision stage. This module is directly connected to the servo driver, and its core function is to convert the geometric calculation results described in embodiments four to six above into specific motor speed commands.

[0092] Specifically, the execution control module integrates an asymmetric speed distribution algorithm. When the stage determination module outputs the "entry transition segment" indicator, the execution control module calls the entry control algorithm to calculate the straight-line travel compensation factor for the rear wheels and sends a deceleration command to the rear wheel drive. When the "corner segment" indicator is output, the curvature radius ratio algorithm is called to send differential commands to both the front and rear wheel drive.

[0093] When the "corner exit transition segment" indicator is output, the corner exit smoothing algorithm is invoked to interpolate and adjust the speed of the front wheels. Through this deep integration of hardware and algorithm, the execution control module can correct the linear speed difference between the front and rear wheels in real time, eliminate kinematic interference, and ultimately achieve precise closed-loop control of the transport vehicle's trajectory.

[0094] It should be understood that the data flow between the above modules is bidirectional and real-time. The parameter acquisition module continuously updates the state vector, and the track selection module and stage determination module update their decision instructions accordingly. The execution control module then feeds back the execution results (such as actual rotational speed) to the parameter acquisition module, forming a complete control closed loop. This system architecture design ensures that the curve adaptive control method of this invention can operate stably in the high-cycle, high-precision handling scenarios of semiconductor production lines.

[0095] In Embodiment Eight, this embodiment provides a computer-readable storage medium. This computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the curve adaptive control method for the semiconductor production line transport process described in any one of Embodiments One to Eight.

[0096] Specifically, the computer-readable storage medium referred to in this embodiment can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the control system of a semiconductor transport vehicle, this storage medium is typically integrated into the vehicle's onboard controller (such as a PLC or embedded industrial computer). For example, the storage medium can be a read-only memory (ROM) used to store the system's boot program and core control algorithm code, ensuring that the curve adaptive control logic of this invention is loaded immediately upon system power-on.

[0097] It can also be random access memory (RAM) used to temporarily store intermediate variables such as the relative position state vector, load weight W, and vibration amplitude A calculated in real time during the control process. It can also be flash memory used to store upgradeable firmware versions so that technicians can update control parameters (such as weighting coefficients α, β, or safety thresholds Wth, Ath) according to production line needs.

[0098] It should be understood that although this embodiment mainly describes the cooperation between the storage medium and the processor, in actual applications, when the processor executes the computer program, it will specifically perform the following steps: reading encoder data through the I / O interface to calculate the relative position state vector, calling the threshold comparison algorithm in the memory to make track optimization decisions, and sending asymmetric speed commands to the servo driver. This process realizes the control of the hardware execution unit by the software logic, thereby completing a complete closed loop from data acquisition, logical judgment to action execution. By solidifying the above method steps into a computer program and storing it in the medium, a standard hardware platform can possess the curve adaptive control function of this invention, reducing hardware modification costs and improving the maintainability and scalability of the system.

[0099] In Implementation Method Nine, taking the OHT (Overhead Traveling Vehicle) system of a 12-inch wafer fab as an example, the curve adaptive control method described in Implementation Methods One through Eight is specifically applied and illustrated. This implementation method aims to verify the actual effectiveness of the method in suppressing micro-vibrations and reducing motor load through specific parameter substitution and process deduction.

[0100] Specifically, in this application scenario, the transport vehicle's geometric parameters are set as follows: wheelbase L = 1.2 meters. The track parameters for the target curve are set as follows: curve center radius R = 5.0 meters, inner track radius r = 4.8 meters (i.e., track width 0.4 meters). The transport vehicle's preset operating speed V0 = 5 meters per second. The control system's preset load safety threshold Wth is 80% of the rated load, and the vibration safety threshold Ath is 0.3G.

[0101] During a complete turn, the transport vehicle executes the following control steps: First, before entering the curve, the system performs a track selection optimization step. The parameter acquisition module reads the load sensor data, showing that the current load weight W is 90% of the rated load (i.e., W>Wth), and simultaneously reads the acceleration sensor data, showing that the current chassis vibration amplitude A is 0.1G (i.e., A≤Ath). Based on the logic in Implementation Method Two, the system determines that the transport vehicle is in a heavy-load stable operating condition and decides to prioritize the outer track of the curve to utilize the larger turning radius and reduce the risk of rollover.

[0102] Subsequently, the front wheels of the transport vehicle enter the curve, and the system enters the curve transition control phase. At this time, the front wheels are located on the inner track of the curve, while the rear wheels remain on the straight track. The system calculates the relative position state vector in real time, assuming the front wheel displacement S0 is 0.5 meters. According to the control strategy of Implementation Method Four, the system calculates the first geometric angle θ1 = S0 / r ≈ 0.104 radians (approximately 5.97 degrees) between the front wheels and the center of the curve.

[0103] Based on the wheelbase L and the current displacement, the system calculates the second geometric angle θ2 formed by the lines connecting the front and rear wheels. According to the geometric relationship, θ2 = arcsin(R×(1-cos(θ1)) / L), and substituting into the numerical calculation, θ2 is approximately 0.052 radians. The system determines the straight-line travel compensation factor for the rear wheels based on the difference between θ1 and θ2. At this time, the front-wheel drive unit speed V1 is set to V0 ×r / R = 5 × 4.8 / 5.0 = 4.8 m / s to match the curvature of the curve. The rear-wheel drive unit speed V2 is then reduced and corrected according to the compensation factor, and V2 ≈ 4.95 m / s is calculated. Through this asymmetrical speed distribution (V1=4.8 m / s, V2=4.95 m / s), the displacement increment of the rear wheels on the straight-line is accurately compensated, eliminating the drag effect caused by the circular motion of the front wheels, thus keeping the vehicle body posture stable during the cornering process.

[0104] Next, the front and rear wheels of the transport vehicle fully enter the curve area, and the system enters curve segment control. According to the strategy in Implementation Method Five, the system controls the front and rear wheel drive units to output speed values ​​corresponding to the curvature of their respective tracks. Since the transport vehicle has chosen the outer track, the front wheel (inner wheel) speed V1 remains at 4.8 m / s, and the rear wheel (outer wheel) speed V2 is adjusted to V0 × (r+L) / R (assuming the rear wheel is located on the outer track with a larger radius; the specific radius needs to be calculated based on the track width), ensuring that the transport vehicle's center of gravity trajectory strictly follows the curve's center. During this stage, both the front and rear wheels perform circular motion, maintaining a stable speed difference and preventing tire slippage and motor overload.

[0105] Finally, the front wheels of the transport vehicle exit the curve and enter the straight track, and the system enters the curve exit transition control. At this time, the front wheels need to smoothly transition from tangential speed to linear speed. According to the strategy in Implementation Method Six, the system calculates the angle between the direction correction angle of the front wheels and the line connecting the rear wheels, and determines the interpolation coefficient. As the rear wheels gradually leave the curve, the interpolation coefficient gradually increases from 0 to 1, the front wheel speed V1 smoothly recovers from 4.8 m / s to 5.0 m / s, and the rear wheel speed V2 remains matched with the curve curvature until it is completely out of the curve. This process eliminates the impact caused by sudden speed changes and prevents the "fishtailing" phenomenon.

[0106] Actual test data verified that, using the curve adaptive control method of this embodiment, the chassis vibration amplitude of the transport vehicle when passing through the aforementioned curves was strictly controlled within 0.25G, lower than the preset safety threshold of 0.3G. Simultaneously, the peak torque of the servo motor was reduced by approximately 15% compared to the traditional forced constant-speed control method. This demonstrates that the method of this invention effectively solves the problems of micro-vibration suppression and motor load optimization in semiconductor handling through real-time calculation of relative position state vectors and asymmetric speed allocation, significantly improving wafer handling yield and equipment lifespan.

[0107] It should be understood that although this specification describes various embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0108] The detailed descriptions listed above are merely specific descriptions of feasible embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. All equivalent embodiments or modifications made without departing from the spirit of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for adaptive control of curves in the handling process of a semiconductor production line, characterized in that, include: The system acquires the transport vehicle's own geometric parameters and the track parameters of the target curve, and obtains the relative position state vector of the transport vehicle in the curve in real time. Based on the relative position state vector and the track parameters, combined with the real-time load status and inertial feedback of the transport vehicle, the optimal selection of the inner or outer track of the curve is made. Based on the selected track and the curve section where the transport vehicle is located, and combined with the relative attitude relationship between the front and rear wheels of the transport vehicle, an asymmetrical speed distribution is performed on the front wheel drive unit and the rear wheel drive unit of the transport vehicle to achieve closed-loop correction of the transport vehicle's motion trajectory.

2. The curve adaptive control method according to claim 1, characterized in that, The step of selecting the optimal inner or outer track for the curve by combining the real-time load status and inertial feedback of the transport vehicle includes: Obtain the current load weight W of the transport vehicle and the real-time vibration amplitude A of the chassis; Set the load safety threshold W th Vibration safety threshold A th ; When W>W th And A≤A th When turning, prioritize using the outer track of the curve to utilize the larger turning radius and reduce the risk of rollover; When W≤W th And A>A th When this is the case, the inner track with less vibration transmission should be selected first to suppress micro-vibrations.

3. The curve adaptive control method according to claim 2, characterized in that, The method of combining the real-time load status and inertial feedback of the transport vehicle to select the optimal inner or outer track for the curve also includes the following steps: When W>W th And A>A th When calculating the comprehensive risk index, R = α⋅W / W th +β⋅A / A th , where α and β are preset weighting coefficients; If R exceeds the preset risk limit, emergency braking will be triggered and track selection will be suspended until W or A falls back to a safe range.

4. The curve adaptive control method according to claim 1, characterized in that, The steps of obtaining the transport vehicle's own geometric parameters and the track parameters of the target curve include, and obtaining the relative position state vector of the transport vehicle in the curve in real time, including: Determine the wheelbase between the front and rear wheels of the transport vehicle, the curvature of the inner track of the curve, the curvature of the center track of the curve, and the preset running speed of the transport vehicle. By monitoring the displacement changes of the front wheel drive unit in real time and combining the geometric constraints of the curve, the relative state vector characterizing the real-time position of the transport vehicle in the curve is calculated.

5. The curve adaptive control method according to claim 4, characterized in that, Based on the relative position state vector and the track parameters, combined with the real-time load state and inertial feedback of the transport vehicle, the steps for optimally selecting the inner or outer track of the curve include: Based on the relative position state vector and the track parameters, the transport vehicle is determined to be in one of the following three phases during curve operation: the entry transition phase, the curve phase, and the exit transition phase: When the relative state vector represents the transport vehicle just entering the curve and there is a difference in track type between the front and rear wheels, it is determined that the transport vehicle is in the curve transition section; When the relative state vector indicates that the transport vehicle is completely within the curve area and the front and rear wheel track types are consistent, it is determined that the transport vehicle is in the curve section; When the relative state vector indicates that the transport vehicle is about to leave the curve and there is a difference in track type between the front and rear wheels, it is determined that the transport vehicle is in the curve exit transition section.

6. The curve adaptive control method according to claim 5, characterized in that, When the transport vehicle is in the transition section before entering a curve, the asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle includes: Based on the relative attitude relationship, calculate the first geometric angle between the front wheel and the currently selected center of the curve. Based on the wheelbase of the transport vehicle and its current displacement, calculate the second geometric angle formed by the lines connecting the front and rear wheels; The straight-line travel compensation factor of the rear wheel is determined based on the difference between the first geometric angle and the second geometric angle. The output speed of the rear-wheel drive unit is reduced and corrected according to the compensation factor, while the speed of the front-wheel drive unit is kept in line with the curvature of the currently selected curve.

7. The curve adaptive control method according to claim 5, characterized in that, When the transport vehicle is on a curve, the asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle includes: Based on the ratio of the curvature radii of the inner track and the center track of the currently selected curve, calculate the theoretical linear velocities of the front and rear wheels under circular motion. The front and rear drive units are controlled to output speed values ​​corresponding to the curvature of their respective tracks, so that the center of gravity trajectory of the transport vehicle fits the center of the curve.

8. The curve adaptive control method according to claim 5, characterized in that, When the transport vehicle is in the exit section of a curve, the asymmetrical speed distribution between the front-wheel drive unit and the rear-wheel drive unit of the transport vehicle includes: Based on the geometric relationship between the current position of the transport vehicle and the exit of the curve, calculate the direction correction angle of the front wheels as they turn from the curve to straight-line travel; Based on the geometric constraint that the rear wheel is still inside the curve, calculate the angle between the line connecting the rear wheel and the center of the currently selected curve. By utilizing the coupling relationship between the direction correction angle and the angle between the connecting lines, the interpolation coefficient for the transition of the front wheel from tangential velocity to linear velocity is determined; The output speed of the front-wheel drive unit is smoothly corrected based on the interpolation coefficients, while maintaining the speed of the rear-wheel drive unit in line with the curvature of the currently selected curve.

9. A curve adaptive control system for the handling process in a semiconductor production line, characterized in that, include: The track selection module is configured to decide whether to select the inner or outer track of the curve based on the real-time load status and inertial feedback of the transport vehicle. The parameter acquisition module is configured to acquire the transport vehicle's own geometric parameters, the track parameters of the target curve, and the real-time relative position state vector; The stage determination module is configured to determine whether the transport vehicle is in the entry-to-curve transition section, the curve section, or the exit-to-curve transition section based on the relative position state vector and the track parameters. The execution control module is configured to perform asymmetric speed distribution and closed-loop correction to the front-wheel drive unit and the rear-wheel drive unit based on the selected track and decision stage.

10. A computer-readable storage medium, characterized in that, When the computer program is executed by the processor, it implements the curve adaptive control method in the semiconductor production line handling process as described in any one of claims 1 to 8.

Images