Variable main beam end beam pin shaft angle self-adaptive control method for triangular workshop

Abstract

The invention discloses a variable main beam end beam pin shaft angle self-adaptive control method for a triangular workshop, and relates to the technical field of intelligent control, and the method comprises the steps: firstly, obtaining the position of a cart in real time, and calculating the theoretical rotation angle of a pin shaft through a non-parallel track geometric mapping model; by collecting motion and load data of the pin shaft, a zero-speed critical index and a geometric internal stress factor are extracted, and a self-adaptive model is constructed to accurately estimate real-time rotation resistance of the pin shaft. And based on the resistance estimated value, the system dynamically adjusts the position ring stiffness of the controller: when it is monitored that the resistance exceeds a safety interval, the stiffness is automatically reduced to present a flexible compliance characteristic, and a self-adaptive correction torque is output. According to the method, the problem that the structural internal stress caused by track geometric errors is difficult to detect through traditional vibration spectrum analysis is solved, and precise planning and flexible protection of the pin shaft are achieved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, specifically to an adaptive control method for the variable main beam end beam pin angle in a triangular workshop. Background Technology

[0002] In specialized manufacturing fields such as aerospace and large steel structures, non-rectangular triangular or trapezoidal workshops are common. Cranes operating in these workshops require their main beam span and end beam pin angles to change in real-time with the trolley's position. This results in the pin mechanism being under constant variable configuration, low-speed heavy load, and complex nonlinear friction conditions. In traditional control systems, the pins are highly susceptible to severe geometric lock-up and internal stress due to angle lag or track geometric errors, leading to pin breakage or wheel wear on the rails.

[0003] To monitor such mechanical failures, vibration sensors are often introduced in the industry for condition sensing. However, because vibration sensors typically output weak analog signals at the millivolt or microampere level, their anti-interference capability is extremely poor. At crane sites, high-frequency electromagnetic noise generated by high-power frequency converters, servo motors, and electromagnetic brakes easily couples into the sensor circuitry, resulting in a very low signal-to-noise ratio. This inherent defect of weak electrical signals being submerged by strong electrical noise makes it difficult for the controller to extract the true mechanical vibration characteristics from the chaotic signals, often leading to false alarms or control failure at critical moments. Existing technologies typically employ a rigid position servo combined with foundation filtering. This method is feasible in rectangular workshops with flat tracks and simple operating conditions, but in triangular workshops with varying spans, the fixed control parameters often cannot balance accuracy and flexibility due to the time-varying characteristics of the main beam stiffness, making it difficult to balance between excessive rigidity leading to breakage and excessive flexibility leading to inaccuracy. Specifically: First, the viscous-slip effect generated by the pin under low speed and heavy load manifests as a high-frequency sawtooth wave in the frequency spectrum. This highly overlaps with the frequency spectrum of electromagnetic interference noise. Conventional frequency spectrum analysis is unable to separate this legitimate fluctuation caused by physical friction mechanism, leading the control system to misjudge normal viscous friction as abnormal vibration and incorrectly reduce speed. Second, it is not sensitive to static internal stress. Frequency spectrum analysis mainly focuses on the dynamically changing AC components, while the structural stress caused by track geometric errors often manifests as DC bias of torque. This static stress cannot be effectively detected by vibration frequency spectrum.

[0004] Therefore, the present invention provides an adaptive control method for the variable main beam end beam pin angle in a triangular workshop. Summary of the Invention

[0005] The purpose of this invention is to provide an adaptive control method for the variable main beam end beam pin angle in a triangular workshop, so as to solve the existing problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an adaptive control method for the variable main beam end beam pin angle in a triangular workshop, comprising the following steps: Step S1: Real-time acquisition of the running position of the trolley on the reference track in the triangular workshop, construction of a non-parallel track geometric mapping model, and calculation of the theoretical rotation angle of the pin corresponding to the current running position; Step S2: Collect the real-time rotational angular velocity, actual rotation angle, and real-time load torque of the pin drive mechanism; Step S3: Construct an adaptive model of the pin shaft rotation resistance state. Extract the zero-velocity critical index and geometric internal stress factor from the collected data and use them as inputs to the adaptive model of the pin shaft rotation resistance state to obtain the real-time rotation resistance estimate of the pin shaft. Step S4: Based on the estimated real-time rotational resistance of the pin shaft, adaptively adjust the position ring stiffness of the pin shaft angle controller; when the estimated real-time rotational resistance of the pin shaft falls outside the preset safe range, reduce the position ring stiffness to exhibit flexible compliance characteristics, and control the pin shaft drive mechanism to output the adaptive correction torque under the current working condition.

[0007] A further improvement of this invention lies in that the non-parallel track geometric mapping model is obtained through an electronic cam trajectory control method. First, a triangular workshop global coordinate system is constructed. The main beam has an active pin and a driven pin at both ends. The reference track is defined as the active pin, and the pin rotation angle is defined as the driven pin. Based on the track geometric equations... Establish a function mapping curve between the position of the active pin and the angle of the driven pin, where Y is the reference side travel distance. The angle between the tracks, The initial bias is set; the controller queries the angle command corresponding to the current position according to the function mapping curve, which is used as the theoretical rotation angle of the pin shaft, and generates a feedforward compensation command by combining the angular velocity feedforward signal obtained by differentiating the function mapping curve.

[0008] A further improvement of this invention lies in the following: the construction process of the adaptive model for the pin shaft rotational resistance state includes: using the real-time load torque, real-time rotational angular velocity, and sign function of the pin shaft drive mechanism as regression input vectors; introducing the zero-velocity critical exponent and geometric internal stress factor as external state variables into the model to establish a dynamic mapping relationship between the state variables and the model hyperparameters; based on the SVR model, during the model inference process, firstly, the insensitive loss band parameter at the current moment is calculated using the current zero-velocity critical exponent, and the penalty factor parameter at the current moment is calculated using the current geometric internal stress factor; using the calculated parameters to perform kernel function mapping and deviation judgment on the input vector, if the prediction deviation is less than the insensitive loss band parameter, it is determined to be operating noise and not updated; otherwise, incremental learning and updating are performed, and finally, the real-time rotational resistance estimate of the pin shaft is output.

[0009] A further improvement of this invention is that the insensitive loss band parameter at the current moment is used to automatically and linearly increase the width of the insensitive loss band when the pin is in a low-speed viscous state, causing the zero-speed critical exponent to increase. This allows the sawtooth-shaped high-frequency jumping samples of torque generated by the viscous-slip effect to fall into the insensitive loss band and be regarded as legitimate noise by the model and ignored. The calculation formula is expressed as the sum of the baseline width and the sensitive adjustment term, whereby the sensitive adjustment term is expressed as the product of the adjustment coefficient and the current zero-speed critical exponent.

[0010] A further improvement of this invention is that the creation process of the penalty factor parameter at the current moment specifically includes: when the track geometry error causes the geometric internal stress factor to increase, the penalty factor is automatically reduced inversely proportionally to reduce the sensitivity and weight of the model to the current moment prediction error. The penalty factor at the current moment is expressed as the baseline penalty factor divided by the denominator, where the denominator is the sum of 1 and the penalty adjustment term, and the penalty adjustment term is the product of the adjustment coefficient and the current geometric internal stress factor.

[0011] A further improvement of this invention is that the zero-velocity critical index is expressed as the real-time rotational angular velocity of the pin. As the velocity approaches zero, the zero-velocity critical exponent approaches 1, indicating a high-risk zone of viscous friction. The formula for calculating the zero-velocity critical exponent is then expressed as: ,in, The preset critical speed threshold for lubrication conditions; Rotation angle based on pin theory Compared with the actual rotation angle Calculate the geometric internal stress factor: This characterizes the degree of forced compression of the main beam onto the pin shaft caused by track unevenness or trolley synchronization error.

[0012] A further improvement of this invention is that the control of the flexible compliance characteristics in step S4 is achieved through a variable gain impedance control strategy, including: setting a normal resistance range. When the estimated real-time rotational resistance of the pin falls within the normal resistance range, the position loop proportional gain of the pin angle controller is maintained. The preset high stiffness value is used; when the estimated real-time rotational resistance of the pin falls outside the normal resistance range, the position loop proportional gain is attenuated in real time according to the following rule: in, Estimate the real-time rotational resistance of the pin shaft. This is the torque limiting constant. This is the decay rate coefficient.

[0013] A further improvement of the present invention is that the method further includes a dual-pin-shaft collaborative verification strategy, configured to calculate the real-time rotational resistance estimates of the active pin and the driven pin respectively; calculate the resistance difference between the two; when the resistance difference exceeds a preset torsion threshold N times, it is determined that the main beam has undergone overall torsion. At this time, the trolley correction command is triggered first to adjust the running speed of the trolley on the reference track side or the inclined track side until the resistance difference falls back to a safe range.

[0014] On the other hand, the present invention provides an adaptive control system for the variable main beam end beam pin angle in a triangular workshop, comprising: The data acquisition module is configured to calculate the theoretical rotation angle of the pin based on the position of the trolley using an electronic cam model; The state perception module is configured to collect the pin's motion state and extract the zero-velocity critical index and geometric internal stress factor. The friction state observation module is configured to construct an adaptive model of the pin shaft rotational resistance state, use the zero-velocity critical exponent to adaptively adjust the model's insensitive loss band to shield viscous noise, use the geometric internal stress factor to adaptively adjust the model's penalty factor to suppress internal stress interference, and output the pin shaft's real-time rotational resistance estimate. The adaptive controller is configured to dynamically adjust the position ring stiffness based on the real-time rotational resistance estimate of the pin, thereby performing flexible angle control.

[0015] A further improvement of the present invention is that the friction state observation module includes a hyperparameter dynamic mapping unit, an adaptive model construction unit, and a deviation decision and update unit; First, the hyperparameter dynamic mapping unit uses specific input operating condition factors to adjust the insensitive loss band and penalty factor in real time. Then, the adaptive model building unit performs regression calculations based on the adjusted parameters. Finally, the deviation decision unit activates incremental learning only when the prediction deviation is greater than the dynamically adjusted insensitive band, based on a specific disturbance shielding strategy. This allows the pin shaft to output a real-time rotational resistance estimate while filtering out viscous noise and internal stress interference.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention first introduces a zero-speed critical index as a specific disturbance feature into the adaptive model of the pin rotation resistance state, and dynamically increases the insensitive loss band width of the model based on this index. This solves the technical problem that traditional control algorithms cannot distinguish between "viscous friction torque sawtooth jump" and "real load sudden change" under low-speed heavy-load conditions of the pin. It achieves effective shielding of high-frequency physical noise under boundary lubrication conditions, avoids high-frequency oscillation caused by misjudging friction noise by the controller, and ensures the stability and control accuracy of the pin when it is moving at extremely low speeds.

[0017] 2. By calculating the deviation between the theoretical angle and the actual angle in real time to extract the geometric internal stress factor, and using this factor to inversely reduce the penalty factor of the SVR model, the algorithm solves the problem that when the track is uneven or the trolley synchronization error causes the main beam to generate forced displacement, the algorithm will misjudge the DC bias torque caused by the elastic deformation of the structure as mechanism jamming or lubrication failure. This achieves accurate separation and estimation of the real friction state under complex geometric constraints, and significantly reduces the false alarm rate of the system.

[0018] 3. By using a variable gain impedance control strategy, the position loop stiffness is reduced in real time according to an inverse proportional law when abnormal resistance is detected. This solves the problem of internal stress surge caused by the forced output of torque to counteract track errors or structural deformation in traditional rigid PID control. It realizes the flexible compliance characteristics of the control system from hard to soft, allowing the pin to generate passive following displacement to release the internal stress of the structure, fundamentally eliminating the mechanical risks of pin breakage and wheel wear on the rail. Attached Figure Description

[0019] Figure 1 This is a flowchart of the adaptive control method for the variable main beam end beam pin angle of the present invention used in a triangular workshop; Figure 2 This is a schematic diagram of the rail-mounted bridge crane in the triangular workshop of the present invention; Figure 3 This is a framework diagram of the adaptive control system for the variable main beam end beam pin angle used in the triangular workshop according to the present invention. Detailed Implementation

[0020] 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.

[0021] The term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three cases: A exists alone, A and B exist simultaneously, and B exists alone.

[0022] Example 1 Figure 1 The flowchart of the adaptive control method for the variable main beam end beam pin angle for a triangular workshop disclosed in this embodiment is shown. The steps are as follows: Step S1: Real-time acquisition of the running position of the trolley on the reference track in the triangular workshop, construction of a non-parallel track geometric mapping model, and calculation of the theoretical rotation angle of the pin corresponding to the current running position; Figure 2This embodiment illustrates a schematic diagram of a track-mounted bridge crane in a triangular workshop. The non-parallel track geometric mapping model, controlled by an electronic cam trajectory, does not rely on sensor feedback to determine the rotation angle of the pivot pin. Instead, it directly establishes a preset mathematical model (geometric equation). and The rigid mapping relationship means that if the large vehicle moves 1mm, the pin shaft will automatically rotate the corresponding small angle according to the curve, thus realizing a priori feedforward control. Establish a two-dimensional Cartesian coordinate system for the triangulation in the motion controller: origin. The starting end of the reference track (straight rail) in the triangular workshop is set; the Y-axis (main shaft) extends along the reference track, representing the travel distance Y of the drive pin (or trolley); the X-axis is perpendicular to the reference track, representing the span direction. The main beam is equipped with a drive pin and a driven pin at both ends, with the reference side track defined as the drive pin and the pin rotation angle defined as the driven pin; based on the track geometry equation... Establish a function mapping curve between the position of the active pin and the angle of the driven pin, where Y is the reference side travel distance. The angle between the tracks, This is the initial bias; Based on the above equations, the theoretical rotation angle of the pin is derived. In trigonometric geometry, the pin angle is usually related to the slope of the tangent to the track. For linear inclined tracks: The controller uses fifth-order polynomial interpolation to fit the discrete points into a continuous and smooth function mapping curve; the controller queries the angle command corresponding to the current position based on the function mapping curve, which is used as the theoretical rotation angle of the pin shaft.

[0023] In this embodiment, the rotation of the pin shaft no longer waits for errors to appear; instead, the pin shaft receives the corresponding acceleration feedforward command simultaneously with the acceleration of the trolley, achieving perfect phase synchronization. Furthermore, because the mapping curve has undergone high-order polynomial fitting, even if the trajectory equation has inflection points, the acceleration change of the pin shaft is continuous, avoiding mechanical shock. At this point, regardless of how the trolley speed changes, the geometric correspondence between the pin shaft angle and the trolley position is always locked by the electronic cam, solving the control problem caused by nonlinear kinematic coupling.

[0024] Step S2: Collect the real-time rotational angular velocity, actual rotation angle, and real-time load torque of the pin drive mechanism; Step S3: Construct an adaptive model of the pin shaft rotation resistance state. Extract the zero-velocity critical index and geometric internal stress factor from the collected data and use them as inputs to the adaptive model of the pin shaft rotation resistance state to obtain the real-time rotation resistance estimate of the pin shaft. In the operation of the triangular workshop, it means that the trolley is moving, and the angle is changing continuously but slowly. When the pin bearing is subjected to tens of tons of gravity, and the trolley is moving at a constant speed, the pin bearing rotates at an extremely slow speed (possibly only a fraction of a second). At these extremely low speeds, the pin is in a state of "boundary lubrication." The metal surface will first "stick," manifested as high static friction, with the motor outputting power but the shaft remaining stationary. Then, the torque exceeds the limit and suddenly slips, manifested as low dynamic friction, with the shaft turning sharply. At this time, the motor current (torque) is not a smooth curve, but rather exhibits sawtooth or step-like fluctuations. The cause of these fluctuations is not a change in load, but rather the fluctuation of the friction coefficient between dynamic and static states.

[0025] Furthermore, the track laying may be uneven or rugged, or the two large vehicles may be slightly out of sync. This manifests as the main beam acting like a giant lever, attempting to "pry open" or "squeeze" the end beams. The resulting force is ultimately concentrated on the pin shaft. The load torque of the pin shaft will include a slowly rising DC component, and this part of the torque is unrelated to the load being lifted, but rather to the internal stress of the structure.

[0026] This invention uses an SVR model to predict the real-time rotational resistance of a pivot pin. However, when "viscous-slip" occurs under low-speed heavy load, the current spikes instantly. At this time, the SVR may mistakenly interpret this as a sudden increase in load or jamming, resulting in an incorrect prediction value. This causes the controller to apply excessive force, causing equipment vibration. Furthermore, seeing a large torque, the model may assume a heavy load is being lifted, thus failing to distinguish between useful load and harmful internal stress.

[0027] The specific process of constructing the adaptive model for the rotational resistance state of the pin shaft includes: Phase 1: Data Input and Preprocessing The process begins with real-time sampling of the pin's operating status. The system first synchronously acquires four sets of raw data as input via the underlying sensor interface: one being the real-time load torque of the pin drive motor. The first factor is the motor current, which is calculated to include friction, inertial force, and various types of noise; the second factor is the real-time rotational angular velocity of the pin. The first is used to determine the speed of movement; the second is the actual rotation angle of the pin. Fourth, the theoretical rotation angle of the pin shaft calculated from the upper-level motion planning module. .

[0028] Phase Two: Extraction of Specific Perturbation Features Before building the model, the system first uses the above input data to calculate two key physical state indicators to quantify the degree of disturbance under specific working conditions.

[0029] First, the system calculates the zero-velocity critical index using real-time rotational angular velocity, expressed as the value of the index when the pin's real-time rotational angular velocity... As the velocity approaches zero, the zero-velocity critical exponent approaches 1, indicating a high-risk zone of viscous friction. The formula for calculating the zero-velocity critical exponent is then expressed as: ,in, The preset critical speed threshold for lubrication conditions; This index reflects whether the pin is in the high-risk zone of stick-slip. When the angular velocity approaches zero or the motion reverses, the index value approaches 1, indicating that it is currently in a boundary lubrication state, and the torque signal will inevitably contain sawtooth-shaped nonlinear jumping noise.

[0030] Simultaneously, the system calculates the absolute value of the difference between the theoretical rotation angle and the actual rotation angle to obtain the geometric internal stress factor; and uses the theoretical rotation angle of the pin shaft... Compared with the actual rotation angle Calculate the geometric internal stress factor: This factor characterizes the degree of forced compression of the main beam onto the pin shaft caused by track unevenness or trolley synchronization errors. It reflects the degree of forced displacement of the main beam due to track errors or trolley asynchrony; the larger the deviation value, the more severe the structural compression on the pin shaft, at which point non-frictional structural internal stresses are mixed into the load torque.

[0031] Phase 3: Adaptive adjustment of model parameters: Before performing regression prediction, the system dynamically adjusts the two core hyperparameters of the SVR model based on the features extracted in the second stage: the insensitive loss bandwidth ( ) and penalty factor ( ).

[0032] The specific adjustment logic is as follows: the system monitors the zero-velocity critical index in real time. When this index increases (indicating that the pin has entered the low-speed viscous region), the system automatically increases the insensitive loss bandwidth according to a preset mapping relationship, expressed as... The fault-tolerant pipe used in the active expansion model encapsulates the jagged torque noise, which is caused by the viscosity effect but has no practical significance, inside the pipe so that it is not regarded as a prediction error.

[0033] Simultaneously, the system monitors the geometric internal stress factor. When this factor increases (indicating severe structural compression), the system automatically reduces the penalty factor, expressed as... ;when When the value is very large, it indicates that the pin is under geometric compression. It contains a large DC bias; if C is large (heavy penalty), the model will be forced to fit this bias, mistakenly believing that the friction has increased; at this time, reducing C reduces the model's confidence in the current error; the model will tend to maintain the original friction coefficient estimate rather than be biased by sudden internal stress; thus avoiding misjudging structural stress as lubrication failure.

[0034] In this way, the model's sensitivity to the current error is reduced, preventing overfitting in an attempt to fit the DC bias torque caused by structural deformation, thus maintaining the model's focus on the true friction characteristics.

[0035] Finally, construct the standard input vector for regression. : .

[0036] Phase 4: Kernel Inference and Denoising Output The system utilizes the insensitive loss bandwidth dynamically adjusted through the above steps ( ) and penalty factor ( ), and perform regression analysis on the current load torque.

[0037] The SVR kernel first uses the Gaussian kernel function (RBF kernel) to process the input vector. Mapping to a high-dimensional space, calculate the initial predicted values: The predicted deviation is then compared with the actual load torque. ; Represents the Lagrange multipliers. Corresponding to the upper boundary constraint, For the lower boundary constraint, b represents the bias term.

[0038] Subsequently, the system performs a deviation judgment: if the deviation falls within the insensitive loss band after dynamic expansion... If the system determines that the current fluctuation belongs to "viscous noise" or "structural disturbance within the allowable range," it will not update the model weights and will directly output the smoothed value from the previous moment or the current prediction baseline value; if the deviation exceeds the insensitive loss band... The system determines that a real resistance anomaly has occurred (such as lubrication failure or foreign object jamming), calculates the loss using the penalty factor, updates the support vector, and corrects the model.

[0039] Finally, the system outputs a real-time estimated rotational resistance of the pivot pin. This is represented as a pure torque signal after physical sensing and noise reduction, stripped of interference from low-speed jitter and geometric internal stress, and transmitted to the subsequent flexible controller as the sole basis for adjusting the position loop stiffness.

[0040] This embodiment addresses the viscous-sliding linear friction disturbances caused by the long-term low-speed, heavy-load state of the pin mechanism during variable-span operation in a triangular workshop, as well as structural internal stress disturbances caused by geometric errors of non-parallel tracks. Unlike traditional algorithms, this embodiment introduces a zero-speed critical index and geometric internal stress factor as prior features for specific scenarios. When the pin is in the boundary lubrication region at zero speed or extremely low speed, the zero-speed critical index dynamically widens the insensitive loss band of the SVR, thereby shielding the sawtooth-like high-frequency noise of torque caused by the viscous effect. When excessive geometric angle deviation is detected, the internal stress factor dynamically reduces the penalty factor, suppressing the interference of geometric internal stress on the friction coefficient estimation. This ensures that even in complex variable-span tribological environments, the algorithm can still accurately extract the true pin running resistance, achieving immunity to false jamming (viscous signals).

[0041] In one possible embodiment, the operator jogs the trolley for millimeter-level positioning, at which point the pin rotation speed is extremely low, for example... The system is located in a severely confined boundary lubrication zone. Existing sensors transmit torque data that fluctuates wildly, like saw teeth; traditional controllers mistake this for a drastic load change, causing fluctuating current output and a clicking sound from the pin. The model in this embodiment detects extremely low speeds... It can spike to 0.95, and the SVR's insensitive band can automatically expand by 3 times. The result is that all jagged fluctuations are... With shielding, the SVR outputs a smooth reference friction curve, the controller maintains a stable output torque, and the pin performs micro-motion quietly and smoothly.

[0042] Step S4: Based on the estimated real-time rotational resistance of the pin shaft, adaptively adjust the position ring stiffness of the pin shaft angle controller; when the estimated real-time rotational resistance of the pin shaft falls outside the preset safety range, reduce the position ring stiffness to exhibit flexible compliance characteristics, control the pin shaft drive mechanism to output an adaptive correction torque that matches the current working condition, so that the actual angle of the pin shaft follows the theoretical rotation angle.

[0043] In the operation of variable span cranes in triangular workshops, pin angle control typically employs high-rigidity position servo control (rigid PID). The core logic of a rigid PID controller is that error is resistance. When the track is uneven, the trolley is out of sync, or there are minor errors in geometric calculations, the actual angle of the pin will deviate from the theoretical angle in a physically unavoidable manner (i.e., geometric resistance).

[0044] Rigid controllers misinterpret deviations caused by structural deformation as control errors, thus outputting a huge current (torque) in an attempt to forcibly correct them. This resistance caused by structural limitations leads to extreme shear stress on the pin bearing, which can cause the pin to break, the pin hole to become elliptical, and even cause the trolley to bite the rails over time.

[0045] To solve the above problems, the conventional engineering method is to use the torque limiting method. That is, a maximum output torque threshold is set, and when the motor torque reaches this value, it is directly cut off.

[0046] Torque limiting is a hard cutoff; when the system enters the limiting state, the position loop effectively fails, and the control mode undergoes a step change. This abrupt change often causes the mechanism to oscillate near the limiting point. Limiting is usually a last resort; before reaching the limiting value, the system maintains high stiffness, meaning the system must withstand enormous stress until it hits the red line, making a smooth preventative retreat impossible.

[0047] Therefore, this invention does not employ a hard cut-off, but rather a soft landing. A variable gain impedance control strategy is used, utilizing the actual rotational drag estimate output from the SVR model as feedback to dynamically adjust the proportional gain of the position loop. The pin drive system can be viewed as a spring; under normal operation, it is a rigid spring (high...). To ensure accuracy; when external pressure (abnormal resistance) is detected, it smoothly transforms into a soft spring (low pressure). This allows it to be compressed or stretched, thereby relieving internal stress. The specific steps are as follows: Step S41: Set the normal resistance range This range is set based on empirical operating data of the pin under good lubrication and without geometric interference (e.g., rated torque). ); Step S42: When the estimated real-time rotational resistance of the pin falls within the normal resistance range. At the same time, maintain the position loop proportional gain of the pin angle controller. Preset high stiffness value When the estimated real-time rotational resistance of the pin falls outside the normal resistance range. At that time, the position loop proportional gain is attenuated in real time according to the following rule: in, Estimate the real-time rotational resistance of the pin shaft. This is the torque limiting constant. The attenuation rate coefficient represents an adjustable sensitivity knob. If the stiffness is set too high, it drops rapidly when resistance is encountered, making the system extremely soft and highly protective (suitable for old, fragile equipment); if... If the stiffness is set to a smaller value, the decrease in stiffness is more gradual, and the system still maintains a certain level of resistance (suitable for highly dynamic equipment).

[0048] pass Excess resistance is extracted; only resistance exceeding the normal range will trigger a decrease in stiffness; this ensures that control performance is not affected within the normal operating range, achieving a seamless transition between normal and abnormal states.

[0049] Through the inverse proportional function However, the nonlinear subtraction method, regardless of resistance... The denominator is always greater than the numerator, therefore It always approaches 0 but never becomes negative. If linear subtraction is used, the gain may become negative when the resistance is too large, leading to positive feedback oscillation (system divergence); the inverse proportional structure mathematically guarantees absolute stability.

[0050] This embodiment reduces the proportional gain of the position loop so that the motor output torque no longer forcibly corrects the angle deviation, allowing the actual angle of the pin to lag behind the theoretical rotation angle within a preset range, thereby generating passive compliant displacement to release the internal stress of the structure.

[0051] In another possible embodiment, foundation settlement occurs at a certain point in the workshop, causing the track to tilt outward by 5mm in a localized area. When a large vehicle passes through this area, the rigid main beam is forcibly stretched. The torque sensor reading suddenly increases by 500 N·m. A traditional PID controller, interpreting the angle as incorrect, frantically increases the current to correct it, resulting in wheel flange wear and even breakage of the pin. This embodiment can detect the increasing deviation between the actual and theoretical angles, raising the... The penalty factor C of SVR is automatically reduced (e.g., from 100 to 10); at the same time Abnormal resistance is identified, triggering step S5 to reduce the position ring stiffness. At this point, the system no longer forcibly corrects the angle, but allows the pin to deviate slightly to accommodate the track's tilt; after passing the settlement zone, it slowly returns to the correct position.

[0052] The method also includes a dual-pin-shaft collaborative verification strategy, configured to calculate the real-time rotational resistance estimates of the active pin and the driven pin respectively; calculate the resistance difference between the two, and when the resistance difference exceeds a preset torsion threshold N times, it is determined that the main beam has undergone overall torsion. At this time, the trolley correction command is triggered first to adjust the running speed of the trolley on the reference track side or the inclined track side until the resistance difference falls back to a safe range.

[0053] The threshold and weight settings involved in this embodiment can be set by default according to the present invention, or can be set by those skilled in the art.

[0054] Example 2 Figure 3This invention presents a framework diagram of a variable main beam end beam pin angle adaptive control system for a triangular workshop. Based on the same inventive concept as Embodiment 1, this invention provides a variable main beam end beam pin angle adaptive control system for a triangular workshop, comprising: The data acquisition module is configured to calculate the theoretical rotation angle of the pin based on the position of the trolley using an electronic cam model; The state perception module is configured to collect the pin's motion state and extract the zero-velocity critical index and geometric internal stress factor. The friction state observation module is configured to construct an adaptive model of the pin shaft rotational resistance state, use the zero-velocity critical exponent to adaptively adjust the model's insensitive loss band to shield viscous noise, use the geometric internal stress factor to adaptively adjust the model's penalty factor to suppress internal stress interference, and output the pin shaft's real-time rotational resistance estimate. The adaptive controller is configured to dynamically adjust the position ring stiffness based on the real-time rotational resistance estimate of the pin, thereby performing flexible angle control.

[0055] The friction state observation module includes a hyperparameter dynamic mapping unit, an adaptive model construction unit, and a deviation decision and update unit. First, the hyperparameter dynamic mapping unit uses the input specific operating condition factors to adjust the insensitive loss band and penalty factor in real time; Subsequently, the adaptive model building unit performs regression calculations using the adjusted parameters; it stores the support vector set and corresponding weights, and receives the input vector; it maps the input to a high-dimensional feature space using the radial basis function, and calculates the preliminary predicted resistance value based on the current support vectors. Finally, based on a specific disturbance shielding strategy, the deviation decision unit activates incremental learning only when the prediction deviation is greater than the dynamically adjusted insensitive band, thereby filtering out viscous noise and internal stress interference while outputting the real-time rotational resistance estimate of the pin.

[0056] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0060] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. An adaptive control method for the variable main beam end beam pin angle in a triangular workshop, characterized in that: Includes the following steps: Step S1: Real-time acquisition of the running position of the trolley on the reference track in the triangular workshop, construction of a non-parallel track geometric mapping model, and calculation of the theoretical rotation angle of the pin corresponding to the current running position; Step S2: Collect the real-time rotational angular velocity, actual rotation angle, and real-time load torque of the pin drive mechanism; Step S3: Construct an adaptive model of the pin shaft rotation resistance state. Extract the zero-velocity critical index and geometric internal stress factor from the collected data and use them as inputs to the adaptive model of the pin shaft rotation resistance state to obtain the real-time rotation resistance estimate of the pin shaft. Step S4: Based on the estimated real-time rotational resistance of the pin shaft, adaptively adjust the position ring stiffness of the pin shaft angle controller; When the estimated real-time rotational resistance of the pin falls outside the preset safety range, the position ring stiffness is reduced to exhibit flexible compliance characteristics, and the pin drive mechanism is controlled to output an adaptive correction torque for the current working condition.

2. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 1, characterized in that: The non-parallel track geometric mapping model is obtained through an electronic cam trajectory control method. First, a triangular workshop global coordinate system is constructed. The main beam has an active pin and a driven pin at both ends. The reference track is defined as the active pin, and the pin rotation angle is defined as the driven pin. Based on the track geometric equations... Establish a function mapping curve between the position of the active pin and the angle of the driven pin, where Y is the reference side travel distance. The angle between the tracks, The initial bias is set; the controller queries the angle command corresponding to the current position according to the function mapping curve, which is used as the theoretical rotation angle of the pin shaft, and generates a feedforward compensation command by combining the angular velocity feedforward signal obtained by differentiating the function mapping curve.

3. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 2, characterized in that: The process of constructing the adaptive model for the rotational resistance state of the pin shaft includes: The real-time load torque, real-time rotational angular velocity, and sign function of the pin drive mechanism are used as regression input vectors. The zero-velocity critical exponent and geometric internal stress factor are introduced into the model as external state variables to establish a dynamic mapping relationship between the state variables and the model hyperparameters. Based on the SVR model, during model inference, the insensitive loss band parameter at the current moment is first calculated using the current zero-velocity critical exponent, and the penalty factor parameter at the current moment is calculated using the current geometric internal stress factor. The calculated parameters are used to perform kernel function mapping and bias judgment on the input vector. If the prediction bias is less than the insensitive loss band parameter, it is determined to be operating noise and will not be updated; otherwise, incremental learning and updating are performed, and finally, the real-time rotational resistance estimate of the pin is output.

4. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 3, characterized in that: The insensitive loss band parameter at the current moment is used to automatically and linearly increase the width of the insensitive loss band when the pin is in a low-speed viscous state, causing the zero-speed critical exponent to increase. This allows the sawtooth-shaped high-frequency jumping samples of torque generated by the viscous-slip effect to fall into the insensitive loss band and be ignored by the model as legitimate noise. The calculation formula is expressed as the sum of the baseline width and the sensitive adjustment term, which is expressed as the product of the adjustment coefficient and the current zero-speed critical exponent.

5. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 3, characterized in that: The process of creating the penalty factor parameter at the current moment specifically includes: when the track geometry error causes the geometric internal stress factor to increase, the penalty factor is automatically reduced inversely proportionally to reduce the sensitivity and weight of the model to the current moment prediction error. The penalty factor at the current moment is expressed as the baseline penalty factor divided by the denominator, where the denominator is the sum of 1 and the penalty adjustment term, and the penalty adjustment term is the product of the adjustment coefficient and the current geometric internal stress factor.

6. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 3, characterized in that: The zero-velocity critical index is expressed as the real-time rotational angular velocity of the pin. As the velocity approaches zero, the zero-velocity critical exponent approaches 1, indicating a high-risk zone of viscous friction. The formula for calculating the zero-velocity critical exponent is then expressed as: ,in, The preset critical speed threshold for lubrication conditions; Rotation angle based on pin theory Compared with the actual rotation angle Calculate the geometric internal stress factor: This characterizes the degree of forced compression of the main beam onto the pin shaft caused by track unevenness or trolley synchronization error.

7. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 1, characterized in that: The control of the flexible compliance characteristics in step S4 is achieved through a variable gain impedance control strategy, including: setting a normal resistance range. When the estimated real-time rotational resistance of the pin falls within the normal resistance range, the position loop proportional gain of the pin angle controller is maintained. The preset high stiffness value is used; when the estimated real-time rotational resistance of the pin falls outside the normal resistance range, the position loop proportional gain is attenuated in real time according to the following rule: in, Estimate the real-time rotational resistance of the pin shaft. This is the torque limiting constant. This is the decay rate coefficient.

8. The adaptive control method for the variable main beam end beam pin angle of a triangular workshop according to claim 1, characterized in that: The method also includes a dual-pin-shaft collaborative verification strategy, configured to calculate the real-time rotational resistance estimates of the active pin and the driven pin respectively; calculate the resistance difference between the two, and when the resistance difference exceeds a preset torsion threshold N times, it is determined that the main beam has undergone overall torsion. At this time, the trolley correction command is triggered first to adjust the running speed of the trolley on the reference track side or the inclined track side until the resistance difference falls back to a safe range.

9. A variable main beam end beam pin angle adaptive control system for a triangular workshop, used to execute the variable main beam end beam pin angle adaptive control method for a triangular workshop as described in any one of claims 1-8, characterized in that: include: The data acquisition module is configured to calculate the theoretical rotation angle of the pin based on the position of the trolley using an electronic cam model; The state perception module is configured to collect the pin's motion state and extract the zero-velocity critical index and geometric internal stress factor. The friction state observation module is configured to construct an adaptive model of the pin shaft rotational resistance state, use the zero-velocity critical exponent to adaptively adjust the model's insensitive loss band to shield viscous noise, use the geometric internal stress factor to adaptively adjust the model's penalty factor to suppress internal stress interference, and output the pin shaft's real-time rotational resistance estimate. The adaptive controller is configured to dynamically adjust the position ring stiffness based on the real-time rotational resistance estimate of the pin, thereby performing flexible angle control.

10. The adaptive control system for the variable main beam end beam pin angle of a triangular workshop according to claim 9, characterized in that: The friction state observation module includes a hyperparameter dynamic mapping unit, an adaptive model construction unit, and a deviation decision and update unit. First, the hyperparameter dynamic mapping unit uses specific input operating condition factors to adjust the insensitive loss band and penalty factor in real time. Then, the adaptive model building unit performs regression calculations based on the adjusted parameters. Finally, the deviation decision unit activates incremental learning only when the prediction deviation is greater than the dynamically adjusted insensitive band, based on a specific disturbance shielding strategy. This allows the pin shaft to output a real-time rotational resistance estimate while filtering out viscous noise and internal stress interference.

Images