Adjusting control method and system for suction cup fork tine feeding and discharging manipulator

Abstract

The application discloses a kind of adjusting control method and system for suction disc fork tooth up and down material manipulator, it is related to automation control technical field, the method is by establishing suction establishment observation window W1 in central control system, and frost layer melting discriminant Cf is calculated in suction establishment observation window W1, and according to preliminary comparison evaluation, the comparison of " hard brittle frost layer dominant stage " and " the dangerous transition interval of melt to water film formation " is realized.By this means, the online identification and stage division of the adsorption interface state are realized, avoiding the misjudgment caused by using only a fixed vacuum threshold or a fixed delay, enabling the manipulator to enter the micro-pulse adsorption defrosting mode and return to S2 when it has not entered the frost layer melting window, thereby reducing the problem of slow or fluctuating vacuum establishment caused by the formation of micro-gaps due to the toppling of frost crystals in the hard brittle frost layer stage, and reducing the instability of handling caused by the falling of frost fragments and adsorption discontinuity.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, specifically to an adjustment and control method and system for a suction cup fork-tooth loading and unloading robot. Background Technology

[0002] With the increasing adoption of automated cold chain warehousing and cryogenic assembly lines, the fields of industrial robot control and automated material handling have placed higher demands on the stable gripping and cycle-time handling of cryogenic sheet metal workpieces. This is particularly true in the operation of suction cup and fork-tooth composite end effectors for loading and unloading robots. A common task is to move insulation boards from a "cold storage buffer pallet" to an "insulation box assembly station," requiring that the entire process of adsorption establishment, initial handling, and placement be free of frost, slippage, and damage, while maintaining a stable cycle time. However, the interface state corresponding to this task exhibits a significant phased abnormality mechanism. In the initial stage, the frost layer has a hard and brittle structure, easily breaking frost crystals and forming micro-gaps upon suction cup contact, leading to a slower or fluctuating vacuum establishment speed. Subsequently, the frost layer partially melts, forming a thin water film, which suddenly enhances the adsorption seal but simultaneously reduces surface friction. This makes the initial stage of handling prone to lateral slippage under lateral acceleration disturbances. Therefore, adjustment and control methods targeting these phased mechanisms are needed to achieve coordinated stability of adsorption and motion.

[0003] Currently, the surface of cryogenic insulation boards is typically covered with a frost layer. In the initial stage of suction cup adsorption, this frost layer has a hard, brittle granular structure, and there is an interface of frost crystals and micro-gaps between the suction cup and the workpiece. This easily leads to localized air leakage, discontinuous adhesion, and fluctuations in the vacuum establishment curve. As the adsorption process continues or localized heat exchange occurs, the frost layer may enter a melting stage and form a water film in the adsorption area. This water film fills the micro-gaps and significantly alters the sealing mechanism, causing abrupt changes in vacuum characteristics. Simultaneously, the water film significantly reduces the interfacial friction coefficient. In existing technologies, suction cup fork-tooth loading and unloading robots often use fixed vacuum thresholds, fixed delays, or fixed speed curves for control. They typically do not distinguish between the hard, brittle frost stage and the melting water film stage, lacking an interfacial state discrimination mechanism based on the coupling characteristics of vacuum changes and surface temperature. This makes it impossible to finely adjust the adsorption establishment rhythm and the initial impact of handling within the frost melting window.

[0004] Because existing control methods struggle to identify the transition window from hard and brittle frost to melting, robotic arms often begin handling directly before the frost layer is effectively broken up or a water film has started to form. This can easily lead to two types of anomalies: First, the hard and brittle frost layer breaks into large pieces during the adsorption and establishment process, causing contamination, affecting subsequent assembly sealing, or triggering false sensor detections. Second, after the melting water film forms, the interfacial friction drops sharply, and the lateral acceleration and rate of change of acceleration in the initial stage can easily trigger relative slippage of the workpiece, causing gripping deviation, collisions, material dropping, or placement errors. In severe cases, this can lead to downtime for cleaning and a decrease in cycle time, reducing the handling stability and operational safety in cold chain scenarios. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an adjustment and control method and system for a suction cup fork-tooth loading and unloading robot, solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution, comprising the following steps:

[0007] S1. By setting up collection points and using the sensor groups set up at the collection points, the operating data is collected in real time and transmitted to the central control system, where an observation window is established.

[0008] S2. Extract features from the running data within the observation window to obtain window feature data. Calculate the frost melting discriminant Cf based on the window feature data, combined with the sampling period Δt and the number of sampling points N1. Perform a preliminary comparative evaluation based on the output of the frost melting discriminant Cf.

[0009] S3. After preliminary comparative evaluation and determination that the frost layer has entered the melting window, the adsorption and motion synergistic limiting control mechanism is implemented, and the synergistic risk quantity Ram is calculated based on the frost layer melting discrimination quantity Cf within the observation window.

[0010] S4. The preset collaborative risk threshold Rth and collaborative risk quantity Ram are used to make a collaborative judgment, and the optimization convergence mechanism and security degradation are executed based on the judgment result.

[0011] Preferably, S1 includes S11;

[0012] S11. By setting several collection points on the suction cup fork loading and unloading robot and setting sensor groups in the collection points, the operation data of the suction cup fork loading and unloading robot during the adsorption and operation process is collected in real time, and the operation data is transmitted to the central control system of the suction cup fork loading and unloading robot through the wireless communication network.

[0013] The data acquisition points include vacuum pressure acquisition points, surface temperature acquisition points, and motion state acquisition points;

[0014] The operational data includes suction cup vacuum pressure sequence, workpiece surface temperature sequence, end-effector acceleration time sequence, and end-effector Z-axis position time sequence;

[0015] Among them, each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence and end Z-direction position time sequence specifically includes the suction cup end vacuum pressure Pv(t) at time t, the workpiece adsorption area surface temperature Ts(t) at time t, the end acceleration a(t) at time t and the end Z-direction position z(t) at time t.

[0016] The sensor group includes a vacuum pressure sensor, an infrared temperature sensor, and a robotic arm controller.

[0017] Preferably, S1 further includes S12;

[0018] S12. The central control system receives operational data in real time and performs data access processing on the operational data through the time synchronization unit of the central control system. The data access processing includes channel mapping of data from different sources, data integrity verification, and unified timestamp calibration. The time synchronization unit unifies the zero time with the adsorption establishment command trigger time and aligns the time axis of each time series in the operational data.

[0019] Among them, when the sampling frequency of the surface temperature time of the workpiece adsorption region is lower than the sampling frequency of the vacuum pressure time, nearest neighbor alignment is used to align the surface temperature time series of the workpiece adsorption region to the sampling timestamp of the vacuum pressure time series.

[0020] After the time axis alignment is completed, an adsorption establishment observation window W1 is established. The start time of the adsorption establishment observation window W1 is the trigger time of the adsorption establishment command. The duration of the adsorption establishment observation window W1 is determined by the preset number of sampling points N1 and the sampling period Δt.

[0021] Then, for each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence, and end Z-axis position time sequence, the suction cup end vacuum pressure Pv(t), workpiece adsorption area surface temperature Ts(t), end Z-axis position z(t), and end acceleration a(t) at time t are subjected to dimensionless processing to eliminate the influence of unit dimensions. The dimensionless processing is performed by normalizing the sampled values ​​based on the baseline mean and fluctuation range of each time sequence within the corresponding window to eliminate the influence of unit dimensions of each parameter.

[0022] Preferably, S2 includes S21;

[0023] S21. Based on the time series of vacuum pressure and the time series of surface temperature of the workpiece adsorption area collected within the adsorption observation window W1; within the adsorption observation window W1, the collection points are divided into a first half sampling point set and a second half sampling point set according to time sequence. The average vacuum pressure Pv,start of the first half sampling point set and the average vacuum pressure Pv,end of the second half sampling point set are calculated respectively. The average surface temperature Ts,start of the first half sampling point set and the average surface temperature Ts,end of the second half sampling point set are also calculated respectively. Based on the sampling period Δt and the number of sampling points N1 of the adsorption observation window W1, the frost layer melting discriminant Cf is calculated to quantitatively analyze the coupling relationship between the intensity of vacuum change and the surface temperature rise trend during the adsorption establishment stage.

[0024] Preferably, S2 further includes S22;

[0025] S22. Based on the output of the frost layer melting discrimination value Cf in the central control system, the central control system performs preliminary comparative evaluation processing to determine the transition stage of the frost layer from hard and brittle to melting; the specific evaluation content is as follows:

[0026] When the frost layer melting discrimination value Cf>1, it indicates that the vacuum change is drastic and the temperature rise is insufficient. It is determined that the workpiece adsorption interface is in a state dominated by hard and brittle frost layer, and the evaluation result of not entering the frost layer melting window is output. The suction cup fork tooth loading and unloading robot is controlled to enter the micro-pulse adsorption and defrost mode and return to S2.

[0027] When the frost layer melting discrimination value Cf≤1, it indicates that the temperature change begins to be significant. The workpiece adsorption interface is determined to have entered the frost layer melting window and the evaluation result of entering the frost layer melting window is output, thereby triggering the adsorption and motion coordinated limiting control mechanism.

[0028] Preferably, S3 includes S31;

[0029] S31. Under the condition of preliminary comparative evaluation of the adsorption and motion coordinated limiting control mechanism, the central control system sends an adsorption establishment slope shaping control command to the vacuum valve control unit.

[0030] The adsorption establishment slope shaping control command shapes the slope of the vacuum establishment process during the adsorption establishment stage, so that the rise rate of the suction cup vacuum pressure sequence during the adsorption establishment period is limited and changes smoothly.

[0031] The central control system sends a short-distance verification segment motion command to the robot controller, so that the suction cup fork-tooth loading and unloading robot can perform the short-distance verification segment motion after completing the pre-contact micro-pressure and entering the adsorption establishment state; the short-distance verification segment motion applies a coordinated amplitude limiting constraint to the motion planning parameters through the central control system, and the coordinated amplitude limiting constraint includes at least the upper limit constraint of lateral acceleration and the upper limit constraint of acceleration change rate.

[0032] The upper limit constraint on lateral acceleration limits the maximum acceleration along the horizontal plane during short-distance verification segment motion.

[0033] The upper limit constraint on the rate of change of acceleration limits the maximum rate of change of acceleration over time during the short-distance verification segment motion.

[0034] Then, based on the starting moment of the short-distance verification segment motion, a verification segment evaluation window W2 is established, and the vacuum pressure time series and the terminal acceleration time series are collected within the verification segment evaluation window W2.

[0035] Preferably, S3 further includes S32;

[0036] S32. Establish a verification segment evaluation window W2 at the start moment of the short-distance verification segment motion, and collect the suction cup vacuum pressure sequence and the end acceleration time sequence within the verification segment evaluation window W2.

[0037] Then, based on the suction cup vacuum pressure Pv(t) and the terminal Z-axis position z(t) at time t of the corresponding sequence point in the suction cup vacuum pressure sequence and the terminal acceleration time sequence, and after dimensionless processing, combined with the frost layer melting discrimination quantity Cf, the cooperative risk quantity Ram is calculated and output to quantitatively analyze the degree of coupling risk between vacuum fluctuation and start-up impact under the frost layer melting window condition.

[0038] Preferably, S4 includes S41;

[0039] S41. Based on the historical collaborative risk quantity Ram, extract the critical value of the coupling risk between vacuum fluctuation and initial impact under the frost melting window condition and set it as the collaborative risk threshold Rth.

[0040] The real-time acquired collaborative risk quantity Ram is then used to make a collaborative judgment with the collaborative risk threshold Rth, and the collaborative judgment result is output. The specific judgment content is as follows:

[0041] When the collaborative risk level Ram ≤ collaborative risk threshold Rth, the verification result is output and the suction cup fork loading and unloading robot is allowed to switch from the short-distance verification section movement to the main transport section movement.

[0042] When the collaborative risk amount Ram is greater than the collaborative risk threshold Rth, the verification failure result is output and the optimization convergence mechanism is triggered.

[0043] Preferably, S4 further includes S42;

[0044] S42. After triggering the optimization convergence mechanism, the central control system obtains the absolute values ​​of the vacuum pressure variance and the terminal acceleration change rate within the verification section evaluation window W2; and uses the dominant decision rule to determine the dominant term, and executes the corresponding optimization convergence mechanism based on the determined dominant term.

[0045] The dominant determination rules are as follows:

[0046] When the variance of vacuum pressure is ≥1.5 × the absolute value of the rate of change of terminal acceleration, it is determined to be the dominant term of vacuum fluctuation, and adsorption-side convergence is performed.

[0047] When the absolute value of the rate of change of terminal acceleration is ≥1.5×vacuum pressure variance, it is determined to be the dominant term of the initial impact and motion-side convergence is performed.

[0048] If none of the inequalities are satisfied, it is determined to be a coupling-dominant term, and both adsorption-side convergence and motion-side convergence are performed simultaneously.

[0049] The adsorption-side convergence is achieved by the central control system adjusting the pulse width modulation slope shaping and stabilization holding time of the vacuum solenoid valve according to the gear position; the specific convergence process is as follows:

[0050] Set the pulse width modulation period to 20ms;

[0051] The effective duty cycle will be reduced by one level from the current level, with each level being 15%.

[0052] Increase the adsorption stability time by one level from the current level, with each level being 80ms;

[0053] The duty cycle is kept constant during the adsorption stabilization period to allow vacuum pressure fluctuations to converge.

[0054] The motion-side convergence is achieved by the central control system performing coordinated amplitude-limiting convergence of motion parameters in the short-distance verification segment according to gear position. The specific convergence content is as follows:

[0055] The upper limit of lateral acceleration is reduced by one level from the current level, with each level being 0.4.

[0056] The upper limit of the rate of change of acceleration is reduced by one level from the current level, with each level being 3 m / s². 3 ;

[0057] Increase the length of the short-distance verification section by one level, with each level being 30mm.

[0058] After optimizing the convergence mechanism, the collaborative judgment result is updated. If the updated collaborative judgment result is still in the judgment result of verification failure more than 3 times, the central control system performs safety degradation processing. The safety degradation processing includes stopping the movement of the main transport section and retracting to a safe height, performing vacuum release and vacuum breaking actions to remove adsorption and prevent dragging, writing the abnormal working conditions and corresponding workstation numbers and ambient temperature ranges into the data recording unit, and selecting manual intervention according to the upper computer strategy.

[0059] An adjustment and control system for a suction cup fork-tooth loading and unloading robot includes an observation window establishment module, a frost layer melting analysis module, a collaborative analysis module, and an optimization convergence and degradation module;

[0060] The observation window establishment module sets up collection points and uses sensor groups set up at the collection points to collect operational data in real time and transmit it to the central control system, where the observation window is established.

[0061] The frost melting analysis module extracts features from the running data within the observation window to obtain window feature data. Based on the window feature data, it calculates the frost melting discriminant Cf by combining the sampling period Δt and the number of sampling points N1 in the window. Based on the output of the frost melting discriminant Cf, it performs a preliminary comparative evaluation.

[0062] The collaborative analysis module, after determining the entry into the frost melting window through preliminary comparative evaluation, executes the adsorption and motion collaborative limiting control mechanism, and calculates the collaborative risk quantity Ram based on the frost melting discrimination quantity Cf within the observation window;

[0063] The optimization convergence and degradation module performs a collaborative determination based on a preset collaborative risk threshold Rth and a collaborative risk amount Ram, and executes an optimization convergence mechanism and a security degradation based on the determination result.

[0064] This invention provides an adjustment and control method and system for a suction cup fork-tooth loading and unloading robot. It has the following beneficial effects:

[0065] (1) This method establishes an adsorption observation window W1 by establishing the vacuum pressure sequence of the suction cup end and the surface temperature sequence of the workpiece adsorption area in the central control system, and extracts the features of the running data in the window according to the sampling point set of the first half and the sampling point set of the second half, and calculates the average vacuum pressure Pv,start of the first half of the window, the average vacuum pressure Pv,end of the second half of the window, the average surface temperature Ts,start of the first half of the window and the average surface temperature Ts,end of the second half of the window, and then calculates the frost layer melting discrimination quantity Cf by combining the sampling period Δt and the number of sampling points N1 of the window, and realizes the comparison between the "hard and brittle frost layer dominant stage" and the "dangerous transition zone from melting to water film formation" based on the preliminary comparative evaluation. This enables online identification and stage division of the adsorption interface state, avoiding misjudgments caused by using only a fixed vacuum threshold or fixed delay. It allows the robot to enter the micro-pulse adsorption and defrosting mode and return to S2 before entering the frost melting window, thereby reducing the problem of slow vacuum establishment or fluctuation caused by micro-gap generated by top-broken frost crystals in the hard and brittle frost stage, and reducing the instability of handling caused by frost fragments falling and adsorption discontinuity.

[0066] (2) After preliminary comparison and evaluation to determine that the frost layer has entered the melting window, the method uses vacuum solenoid valve pulse width modulation to perform adsorption establishment slope shaping control, so that the vacuum pressure sequence of the suction cup during adsorption establishment is limited and changes smoothly; at the same time, short-distance verification segment movement is performed, and the upper limit of lateral acceleration and the upper limit of acceleration change rate are synergistically limited, thereby reducing the probability of lateral slippage induced by starting impact under low friction water film conditions. Within the evaluation window W2 of the verification section, the variance of vacuum pressure and the absolute value of the rate of change of terminal acceleration are further calculated, and a collaborative risk quantity Ram is constructed. Then, Ram is collaboratively judged with the collaborative risk threshold Rth extracted based on historical Ram. When the optimization convergence mechanism is triggered, adsorption-side convergence and motion-side convergence are executed respectively according to the judgment rules of "vacuum fluctuation-dominated, start-up impact-dominated, or coupling-dominated". The duty cycle, stable holding time, upper limit of lateral acceleration, upper limit of rate of change of acceleration, and length of verification section are clearly adjusted in a gear manner to achieve targeted suppression of vacuum fluctuation and start-up impact, thereby improving the throughput of verification section and the stability of main transport section, and reducing the risk of material drop, collision and workpiece breakage.

[0067] (3) This method diverts the control strategy by using the preliminary comparative evaluation results of the frost layer melting discrimination quantity Cf, so that the "entering the frost layer melting window" judgment result output by innovation point 1 directly triggers the adsorption and motion synergistic limiting control mechanism of innovation point 2, forming a closed-loop control chain of "state recognition, strategy triggering, verification evaluation, parameter convergence, and safety degradation": when Cf>1, it enters the micro-pulse adsorption defrosting mode and returns to S2 to improve the interface adhesion of the hard and brittle frost layer stage; when Cf≤1, it starts the slope shaping and synergistic limiting and calculates Ram in W2 to quantify the risk, and then realizes parameter adaptive optimization through Rth threshold judgment and the gear convergence mechanism of S42, and writes the effective gear into the data recording unit for subsequent strategy library updates. Therefore, the method can take into account the two different mechanisms of frost layer breakage and water film slippage under the same set of operational data acquisition and window analysis framework. It not only improves the stability and consistency of cold chain handling tasks, but also avoids dragging and damage through safety degradation when continuous optimization still fails, thus achieving a comprehensive and stable handling effect without frost shedding, slippage and damage. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the adjustment and control method for a suction cup fork-tooth loading and unloading robot according to the present invention.

[0069] Figure 2 This is a schematic diagram of the adjustment and control system for a suction cup fork-tooth loading and unloading robot according to the present invention.

[0070] Figure 3 Top view of the mechanical structure of the suction cup fork-tooth loading and unloading robot;

[0071] Figure 4 This is a side view of the mechanical structure of a suction cup fork-tooth loading and unloading robot. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] Example 1

[0074] Please see Figure 1 This invention provides an adjustment and control method for a suction cup fork-tooth loading and unloading robot. To achieve the above objectives, this invention is implemented through the following technical solution, including the following steps:

[0075] S1. By setting up collection points and using the sensor groups set up at the collection points, the operating data is collected in real time and transmitted to the central control system, where an observation window is established.

[0076] S2. Extract features from the running data within the observation window to obtain window feature data. Calculate the frost melting discriminant Cf based on the window feature data, combined with the sampling period Δt and the number of sampling points N1. Perform a preliminary comparative evaluation based on the output of the frost melting discriminant Cf.

[0077] S3. After preliminary comparative evaluation and determination that the frost layer has entered the melting window, the adsorption and motion synergistic limiting control mechanism is implemented, and the synergistic risk quantity Ram is calculated based on the frost layer melting discrimination quantity Cf within the observation window.

[0078] S4. The preset collaborative risk threshold Rth and collaborative risk quantity Ram are used to make a collaborative judgment, and the optimization convergence mechanism and security degradation are executed based on the judgment result.

[0079] In this embodiment, the method arranges vacuum pressure acquisition points, surface temperature acquisition points, and motion state acquisition points in the vacuum branch / combination cavity of the suction cup, the field of view of the workpiece adsorption area, and the robot controller, respectively. Pv(t), Ts(t), a(t), and z(t) are collected in real time and transmitted to the central control system to establish observation windows W1 and W2. The purpose is to observe "adsorption interface sealing changes" and "start-up impact disturbances" on the same time axis. The physical significance of this setting is that the frost layer state changes rapidly and in stages under cold storage conditions. If only a single-point vacuum threshold is used, it is easy to misjudge "air leakage fluctuations caused by the breakage of hard and brittle frost crystals" as "insufficient adsorption", which will lead to unstable cycle or frequent false stops. S2 extracts features from the mean values ​​of the first and second halves of W1 and calculates Cf. It uses Cf > 1 and Cf ≤ 1 to achieve rapid flow separation in the "hard and brittle frost layer stage / melting water film transition window". The main problem in the hard and brittle stage is that the frost crystals break and form micro gaps, which leads to slow or fluctuating vacuum establishment. Although the sealing will suddenly be enhanced in the melting stage, the water film will cause the friction to drop sharply, making it very easy to slip laterally at the start. Therefore, using Cf as the trigger condition can avoid two types of miscontrol: "decelerating before entering the melting window" or "starting normally after entering the melting window". After determining that it has entered the melting window, S3 uses a vacuum solenoid valve slope shaping to make the vacuum rise a smooth process. At the same time, it limits the lateral acceleration and the rate of change of acceleration within a short verification section (e.g., 50–80 mm) and calculates Ram in W2. The purpose is to limit the most dangerous "start-up transient" to a short verification distance: if the water film causes low friction, excessive lateral acceleration or impact will directly trigger slippage, while vacuum fluctuations will reflect unstable interface sealing. Ram quantifies the coupling risk of the two before reliable release can be made. S4 determines the Ram and Rth thresholds and performs graded convergence according to the dominant factor (if vacuum fluctuation is dominant, the duty cycle is lowered and the holding time is extended; if the starting impact is dominant, the upper limit of the rate of change is tightened, the upper limit of the lateral acceleration is reduced, and the verification section is lengthened). If the upper limit of the number of times is exceeded, a safety downgrade of vacuum breaking is performed by reversing. The direct effect of this implementation process is to reduce the two types of risks of "frost falling (frost crystal breakage)" and "slippage (low friction of water film)" without changing the hardware. This makes the cycle time of handling from the cold storage buffer pallet to the insulated box assembly station more stable, the probability of material falling and collision is lower, and the surface damage and position deviation of the workpiece are significantly reduced.

[0080] Example 2

[0081] Please see Figure 1 , Figure 3 and Figure 4 Specifically: S1 includes S11;

[0082] S11. By setting several collection points on the suction cup fork loading and unloading robot and setting sensor groups in the collection points, the operation data of the suction cup fork loading and unloading robot during the adsorption and operation process is collected in real time, and the operation data is transmitted to the central control system of the suction cup fork loading and unloading robot through the wireless communication network.

[0083] The data acquisition points include vacuum pressure acquisition points, surface temperature acquisition points, and motion status acquisition points;

[0084] The operational data includes the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end-effector acceleration time sequence, and end-effector Z-axis position time sequence;

[0085] Among them, each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence and end Z-direction position time sequence specifically includes the suction cup end vacuum pressure Pv(t) at time t, the workpiece adsorption area surface temperature Ts(t) at time t, the end acceleration a(t) at time t and the end Z-direction position z(t) at time t.

[0086] The sensor group includes a vacuum pressure sensor, an infrared temperature sensor, and a robotic arm controller;

[0087] Specifically: A vacuum pressure acquisition point is set in the vacuum branch of the suction cup or the manifold of the suction cup, and a vacuum pressure sensor is set at the vacuum pressure acquisition point to obtain the vacuum pressure Pv(t) at the suction cup end at time t.

[0088] Surface temperature acquisition points are set on the side of the gripping station towards the adsorption area or on the support of the mechanical wrist. Infrared temperature measurement modules are set on the surface temperature acquisition points to obtain the surface temperature Ts(t) of the workpiece adsorption area at time t.

[0089] Motion state acquisition points are set inside the robot controller. These points are used to acquire the Z-axis position z(t) of the end effector at time t and the end effector acceleration a(t) at time t.

[0090] The operational data includes sampled values, sampling period, and sampling timestamp information.

[0091] The output signal of the vacuum pressure sensor is transmitted to the central control system via the controller I / O module using the IO-Link / RS485 digital bus.

[0092] The output signal of the infrared temperature sensor is transmitted to the central control system via a serial port.

[0093] The end Z-axis position z(t) and end acceleration a(t) at time t are transmitted to the central control system by the robot controller through the internal shared data area or real-time industrial Ethernet bus. This enables the central control system to obtain the suction cup end vacuum pressure Pv(t) and workpiece adsorption area surface temperature Ts(t) at time t for interface state characterization, and the end Z-axis position z(t) and end acceleration a(t) at time t for control response characterization.

[0094] S1 also includes S12;

[0095] S12. The central control system receives operational data in real time and performs data access processing on the operational data through the time synchronization unit of the central control system. The data access processing includes channel mapping of data from different sources, data integrity verification, and unified timestamp calibration. The time synchronization unit uses the adsorption establishment command trigger time to unify the zero time and aligns the time axis of each time series in the operational data.

[0096] Among them, when the sampling frequency of the surface temperature time of the workpiece adsorption region is lower than the sampling frequency of the vacuum pressure time, nearest neighbor alignment is used to align the surface temperature time series of the workpiece adsorption region to the sampling timestamp of the vacuum pressure time series.

[0097] After the time axis alignment is completed, an adsorption establishment observation window W1 is established. The start time of the adsorption establishment observation window W1 is the adsorption establishment command trigger time. The duration of the adsorption establishment observation window W1 is determined by the preset number of sampling points N1 and the sampling period Δt.

[0098] Then, for each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence, and end Z-axis position time sequence, the suction cup end vacuum pressure Pv(t), workpiece adsorption area surface temperature Ts(t), end Z-axis position z(t), and end acceleration a(t) at time t are subjected to dimensionless processing to eliminate the influence of unit dimensions. The dimensionless processing is performed by normalizing the sampled values ​​based on the baseline mean and fluctuation range of each time sequence within the corresponding window to eliminate the influence of unit dimensions of each parameter.

[0099] In this embodiment, the method acquires Pv(t), Ts(t), a(t), and z(t) by arranging vacuum pressure acquisition points in the vacuum branch or manifold of the suction cup, surface temperature acquisition points on the gripping station side or wrist support, and motion state acquisition points inside the robot controller. These data are then aggregated to the central control system via IO-Link / RS485, serial port, and the controller's internal shared data area / industrial Ethernet. The purpose is to simultaneously "see" the sealing state of the adsorption interface and the dynamic disturbance of the robot: for example, the vacuum pressure acquisition point being close to the suction cup end can reduce the response lag caused by the pipeline volume, avoiding the misjudgment of "the end has been bonded but the pipeline pressure has not yet been transmitted to the upstream" as insufficient adsorption; the surface temperature acquisition point being aligned with the adsorption area can capture the temperature rise trend of the frost layer from hard and brittle to melting, avoiding the difficulty in distinguishing between "frost crystal top breakage and leakage" and "sudden change in sealing due to water film formation" based solely on the vacuum signal; the motion state acquisition points provide a(t) and z(t), which can reflect the starting impact and bonding stroke, avoiding the misjudgment of interface leakage due to vacuum fluctuations caused by the starting impact. Step S12 performs channel mapping, integrity verification, and unified timestamp calibration on the multi-source data in the central control system. The time axis alignment is completed using the adsorption establishment command trigger time as the unified zero time. When the temperature sampling frequency is lower than the vacuum sampling frequency, nearest neighbor alignment is used to ensure that Pv(t) and Ts(t) are comparable at the same time. Then, the start and end of W1 are determined by the preset N1 and Δt to ensure that feature extraction occurs in the "most sensitive short window of adsorption establishment" rather than the steady state segment, reducing the impact of noise and hysteresis on the judgment. Finally, dimensionless processing of baseline mean and fluctuation range normalization is performed on Pv(t), Ts(t), a(t), and z(t) to ensure that different physical quantities are not dominated by a "larger value" in the subsequent evaluation due to differences in unit scale. By setting up the above, the risks of "air leakage fluctuations caused by frost layer breakage", "water film melting risk caused by temperature rise" and "interface disturbances caused by start-up impact" can be accurately aligned in time and can be quantified and compared. This reduces misjudgments, improves the consistency of judgment, and makes subsequent window feature calculation and control decisions more stable and repeatable, thereby improving the success rate of adsorption establishment, reducing the risk of frost shedding and slippage, and improving cycle stability.

[0100] Example 3

[0101] Please see Figure 1 Specifically: S2 includes S21;

[0102] S21. Based on the time series of vacuum pressure and the time series of surface temperature of the workpiece adsorption area collected within the adsorption observation window W1; within the adsorption observation window W1, the collection points are divided into a first half sampling point set and a second half sampling point set according to time sequence. The average vacuum pressure Pv,start of the first half sampling point set and the average vacuum pressure Pv,end of the second half sampling point set are calculated respectively. The average surface temperature Ts,start of the first half sampling point set and the average surface temperature Ts,end of the second half sampling point set are also calculated respectively. Based on the sampling period Δt and the number of sampling points N1 in the adsorption observation window W1, the frost layer melting discriminant Cf is calculated to quantitatively analyze the coupling relationship between the intensity of vacuum change and the surface temperature rise trend during the adsorption establishment stage.

[0103] The frost melting discriminant Cf is calculated using the following algorithm formula:

[0104] ;

[0105] In the formula, This represents the average rate of change in vacuum. The significance of this setting is that it reflects the speed of change in vacuum intensity during adsorption establishment. In the hard, brittle frost layer stage, the suction cup contacts the interface of "frost crystals + micro-gaps": frost crystals break, localized leakage occurs, and the seal is discontinuous; vacuum establishment often exhibits drastic changes, fluctuations, and sluggish establishment. Using "the difference between the average values ​​of the first and second halves / time" captures the characteristics of the "overall change amplitude" (more noise-resistant than a single point). In the water film melting stage, the water film begins to form and fill the micro-gaps: sealing is enhanced, and vacuum establishment becomes smoother and more predictable; molecules often no longer exhibit the "abnormally drastic" change characteristics of the hard, brittle stage; therefore, the molecular characterization describes the "vacuum establishment dynamics caused by interface sealing."

[0106] This represents the average rate of temperature change; the significance of this setting is: the strength of the thermal drive in the frost melting process; the frost layer must satisfy the trend of "the temperature of the adsorption region rising and approaching the melting condition" from hard and brittle to melting.

[0107] When melting begins, the rate of change in surface temperature is usually more pronounced (even if the overall temperature rise is not large, the trend will be more stable).

[0108] The "1+" before the denominator has two layers of physical / engineering significance:

[0109] Avoid division by zero and numerical explosion: In cold storage environments, the average rate of temperature change is often very small or even close to zero. Directly dividing by zero can lead to unstable judgments.

[0110] Treating temperature rise as a "suppression term": The more significant the temperature rise, the more likely melting is to occur, and the discriminant should be more inclined to "enter the melting window". Therefore, let it play the role of "lowering Cf" in the denominator.

[0111] The denominator describes the "strength of thermal evidence for the occurrence of melting".

[0112] S2 also includes S22;

[0113] S22. Based on the output of the frost layer melting discrimination value Cf in the central control system, the central control system performs preliminary comparative evaluation processing to determine the transition stage of the frost layer from hard and brittle to melting; the specific evaluation content is as follows:

[0114] When the frost layer melting discrimination value Cf>1, it indicates that the vacuum change is drastic and the temperature rise is insufficient. The frost layer is still in the "hard and brittle and breakage-dominated stage". It is determined that the workpiece adsorption interface is in the hard and brittle frost layer-dominated state, and the evaluation result of not entering the frost layer melting window is output. The suction cup fork tooth loading and unloading robot is controlled to enter the micro-pulse adsorption and defrost mode and return to S2.

[0115] When the frost layer melting discrimination value Cf≤1, it indicates that the temperature change begins to be significant and the frost layer enters the "dangerous transition zone from melting to water film formation". The workpiece adsorption interface is determined to have entered the frost layer melting window and the evaluation result of entering the frost layer melting window is output, thereby triggering the adsorption and motion coordinated limiting control mechanism.

[0116] The significance of this assessment is to determine whether the interface between the suction cup and the workpiece has entered the transition window of "frost melting and water film formation", thereby deciding whether it is necessary to activate the coordinated limiting control of adsorption and motion to prevent low-friction slippage.

[0117] In this embodiment, step S21 of the method uses the adsorption observation window W1 as the analysis object. The vacuum pressure time series and the surface temperature time series of the workpiece adsorption region within W1 are divided into a first half and a second half according to time sequence. Pv,start, Pv,end, Ts,start, and Ts,end are calculated respectively. The average change rate of vacuum and the average change rate of temperature are characterized by "the difference between the mean values ​​of the first and second halves / window duration (Δt·N1)". Then, the frost layer melting discriminant Cf is calculated. The purpose is not to look at the vacuum size at a certain moment. Instead, it focuses on the overall trend of "whether the sealing mechanism changes during the adsorption establishment process": for example, the crushing of hard and brittle frost crystals will introduce micro gaps, and vacuum establishment often manifests as fluctuations or lags. The mean difference can avoid misjudgments caused by pump pulsation and transient noise. When the frost layer begins to melt, the temperature rise trend is more stable. Even if the temperature rise is not large, it can be reflected by the average rate of change. At the same time, the denominator "1+" avoids the temperature change approaching 0 in low-temperature environments, which would lead to numerical explosion. The temperature rise is used as a suppression term, so that the more obvious the temperature rise, the lower the Cf, and the more likely it is to be judged as entering the melting window. S22 further compares and evaluates Cf with the threshold 1: when Cf>1, it is judged as the stage dominated by hard and brittle crushing and enters the micro-pulse adsorption to defrost and returns to S2. This avoids rashly entering the transportation in the state of "unresolved leakage" which would lead to poor adsorption and frost shedding. When Cf≤1, it is judged as entering the dangerous transition zone from melting to water film formation and triggers the subsequent adsorption and motion synergistic limitation. This avoids the water film causing a sudden drop in friction and the resulting lateral slippage. This "trend discrimination + threshold diversion" implementation method enables online identification and strategy switching of frost layers in stages, reduces confusion between the hard and brittle stage and the melting stage of the vacuum threshold method, thereby improving the consistency of interface judgment, reducing the risk of frost shedding and slippage, and providing a more reliable triggering time and more stable beat performance for subsequent collaborative limiting.

[0118] Example 4

[0119] Please see Figure 1 Specifically: S3 includes S31;

[0120] S31. Under the condition of preliminary comparative evaluation of the adsorption and motion coordinated limiting control mechanism, the central control system sends an adsorption establishment slope shaping control command to the vacuum valve control unit.

[0121] The adsorption establishment slope shaping control command shapes the slope of the vacuum establishment process during the adsorption establishment phase, limiting the rate of increase of the suction cup vacuum pressure sequence during adsorption establishment and making it change smoothly. Slope shaping is achieved by periodically turning the vacuum solenoid valve on and off according to a preset modulation cycle at the beginning of adsorption establishment, gradually transitioning the effective duty cycle of the vacuum solenoid valve from a first duty cycle to a second duty cycle. This limits the rate of increase of the suction cup vacuum pressure sequence and makes it change smoothly. The first duty cycle is smaller than the second duty cycle, and the second duty cycle corresponds to the vacuum solenoid valve entering a continuously conducting state or near-continuously conducting state, to achieve stable adsorption after slope shaping. The central control system monitors the changes in the suction cup vacuum pressure sequence in real time during slope shaping. When the suction cup vacuum pressure sequence reaches the preset stability condition, the pulse width modulation ends and switches to continuous conduction to complete adsorption establishment.

[0122] The central control system sends a short-distance verification segment motion command to the robot controller, so that the suction cup fork-tooth loading and unloading robot can perform short-distance verification segment motion after completing the pre-contact micro-pressure and entering the adsorption establishment state. The short-distance verification segment motion applies coordinated amplitude limiting constraints to the motion planning parameters through the central control system. The coordinated amplitude limiting constraints include at least the upper limit constraint of lateral acceleration and the upper limit constraint of acceleration change rate.

[0123] The upper limit constraint on lateral acceleration limits the maximum acceleration along the horizontal plane during short-distance verification segment motion.

[0124] The upper limit constraint on the rate of change of acceleration limits the maximum rate of change of acceleration over time during the short-distance verification segment motion, enabling the short-distance verification segment motion to be completed under low-impact conditions and providing a stable observation process for subsequent risk assessment.

[0125] Then, based on the starting moment of the short-distance verification segment motion, a verification segment evaluation window W2 is established, and vacuum pressure time series and terminal acceleration time series are collected within the verification segment evaluation window W2.

[0126] S3 also includes S32;

[0127] S32. Establish a verification segment evaluation window W2 at the start moment of the short-distance verification segment motion, and collect the suction cup vacuum pressure sequence and end acceleration time sequence within the verification segment evaluation window W2.

[0128] Then, based on the vacuum pressure Pv(t) at the suction cup end and the Z-axis position z(t) at the end of the suction cup at time t corresponding to the sequence points in the suction cup vacuum pressure sequence and the end acceleration time sequence, and after dimensionless processing, combined with the frost layer melting discrimination quantity Cf, the collaborative risk quantity Ram is calculated and output to quantitatively analyze the degree of coupling risk between vacuum fluctuation and start-up impact under the frost layer melting window condition.

[0129] The collaborative risk quantity Ram is calculated and output using the following algorithm formula:

[0130] In the formula, d represents the differential symbol, dt represents the differential with respect to the time variable, and Var represents the variance.

[0131] This indicates the absolute value of the rate of change of the end-effector acceleration within the evaluation window W2 of the verification section; its physical indication is whether there are any minor changes at the sealing interface between the suction cup and the workpiece during the verification section.

[0132] Why use variance? Variance measures “fluctuation energy” and can quantify the overall pressure fluctuations caused by pump pulsation, frost layer breakage, and micro-slippage.

[0133] Risk Explanation:

[0134] Large variance and shaking of the seal indicate air leakage / adhesion changes / water film slippage, increasing the risk of material falling off / displacement;

[0135] Small variance and stable sealing facilitate entry into the main transport section;

[0136] It represents the variance of vacuum pressure within the evaluation window W2 of the verification section; its physical orientation is the "impact intensity" of the motion initiation, that is, the rate of change of acceleration (often referred to as the jerk's characteristic quantity).

[0137] Why use this amount: Under low-friction water film conditions, slippage is not determined by the velocity itself, but is often triggered by transient impacts; the greater the impact, the easier it is for the interface to experience relative displacement or seal failure.

[0138] Risk Explanation:

[0139] With a high rate of change and a strong initial impact, it is easier to push the water film interface to a sliding state or break up the frost layer.

[0140] With a small rate of change, the start-up is smoother, and the low-friction interface is less prone to instability.

[0141] In this embodiment, after the initial comparative evaluation determines that the frost layer has entered the melting window, step S31 of the method first performs "slope shaping" on the adsorption establishment process. The central control system periodically turns the vacuum solenoid valve on / off with a preset modulation cycle, so that the effective duty cycle gradually transitions from the first duty cycle to the second duty cycle and finally switches to continuous conduction. The purpose is to transform the sudden change in vacuum establishment caused by "sudden enhancement of sealing after water film formation" into a controllable and smooth rise. Because the interface friction decreases during the melting stage, if the vacuum rises suddenly and is superimposed with the impact of the robot starting, transient relative slippage and positional shift are likely to occur. Slope shaping can reduce the impactful changes in adsorption force establishment and reduce the disturbance to the low-friction interface. Subsequently, the robotic arm performs a short-distance verification segment (e.g., 50–80 mm) and coordinates the upper limit of lateral acceleration and the upper limit of the rate of change of acceleration. The purpose is to limit the most sensitive "start-up phase" to short-distance, low-impact conditions for verification: excessive lateral acceleration will directly drive lateral slippage, while excessive rate of change of acceleration will damage the seal stability or trigger water film slippage with transient impact; limiting the amplitude can significantly suppress start-up impact and lateral inertia. Then, W2 is established at the start of the verification segment, and vacuum pressure and end-effector acceleration are collected. In step S32, the coordinated risk quantity Ram is calculated within W2. The variance of vacuum pressure is used as a quantitative indicator of "whether the seal is shaking", and the absolute value of the rate of change of end-effector acceleration is used as a quantitative indicator of "strength of start-up impact". Combined with Cf, a coupled risk characterization is formed, thereby incorporating "interface instability" and "motion disturbance" into the same risk scale. The above implementation method can avoid slippage and material dropping caused by "excessive vacuum establishment + excessive initial impact" after entering the melting window. At the same time, it can expose the sealing fluctuation and impact risk in advance within a short distance, making the subsequent threshold judgment and parameter convergence more based on the evidence. This can improve the stability of the main transport section, reduce the probability of offset collision and workpiece breakage, and improve the cycle consistency.

[0142] Example 5

[0143] Please see Figure 1 Specifically: S4 includes S41;

[0144] S41. Based on the historical collaborative risk quantity Ram, extract the critical value of the coupling risk between vacuum fluctuation and initial impact under the frost melting window condition and set it as the collaborative risk threshold Rth; specifically, it is the critical value of vacuum stability within the melting window.

[0145] The real-time acquired collaborative risk quantity Ram is then used to make a collaborative judgment with the collaborative risk threshold Rth, and the collaborative judgment result is output. The specific judgment content is as follows:

[0146] When the collaborative risk level Ram ≤ collaborative risk threshold Rth, the verification result is output and the suction cup fork loading and unloading robot is allowed to switch from the short-distance verification section movement to the main transport section movement.

[0147] When the collaborative risk amount Ram is greater than the collaborative risk threshold Rth, the verification failure result is output and the optimization convergence mechanism is triggered.

[0148] S4 also includes S42;

[0149] S42. After triggering the optimization convergence mechanism, the central control system obtains the absolute values ​​of the vacuum pressure variance and the terminal acceleration change rate within the verification section evaluation window W2; and uses the dominant decision rule to determine the dominant term, and executes the corresponding optimization convergence mechanism based on the determined dominant term.

[0150] The dominant determination rules are as follows:

[0151] When the variance of vacuum pressure is ≥1.5 × the absolute value of the rate of change of terminal acceleration, it is determined to be the dominant term of vacuum fluctuation, and adsorption-side convergence is performed.

[0152] When the absolute value of the rate of change of terminal acceleration is ≥1.5×vacuum pressure variance, it is determined to be the dominant term of the initial impact and motion-side convergence is performed.

[0153] If none of the inequalities are satisfied, it is determined to be a coupling-dominant term, and both adsorption-side convergence and motion-side convergence are performed simultaneously.

[0154] Note: The 1.5 times here is a specific threshold, which is used to avoid repeatedly switching the convergence direction when two points are close;

[0155] The adsorption-side convergence is achieved by the central control system adjusting the pulse width modulation, slope shaping, and stabilization holding time of the vacuum solenoid valve according to the specified gear position; the specific convergence process is as follows:

[0156] Set the pulse width modulation period to 20ms;

[0157] The effective duty cycle is reduced by one level from the current level, with each level being 15%, for example, from 60% to 45%, in order to reduce the vacuum build-up rate.

[0158] Increase the adsorption stability time by one level from the current level, with each level being 80ms, for example, increasing it from 120ms to 200ms;

[0159] The duty cycle is kept constant during the adsorption stability period to allow vacuum pressure fluctuations to converge.

[0160] The specific meaning of enhancing slope shaping is: reducing the duty cycle (slower and smoother vacuum rise);

[0161] The specific meaning of "extending the adsorption stability retention time" is: the retention time is increased step by step from 120→200→280ms;

[0162] Motion-side convergence is achieved through coordinated amplitude-limiting convergence of motion parameters in the short-distance verification segment by the central control system according to gear position. The specific convergence content is as follows:

[0163] The upper limit of lateral acceleration is reduced by one level from the current level, with each level being 0.4, for example, from 1.2 m / s² to 0.8 m / s²;

[0164] The upper limit of the rate of change of acceleration is reduced by one level from the current level, with each level being 3 m / s². 3 For example, 8m / s 3 Reduced to 5m / s 3 or from 5m / s 3 Reduced to 2m / s 3 ;

[0165] Increase the length of the short-distance verification section by one level from the current level, with each level being 30mm, for example, from 50mm to 80mm, or from 80mm to 110mm;

[0166] The specific meaning of "reduction" is: the upper limit of acceleration and the upper limit of rate of change are reduced step by step according to the gear.

[0167] The specific meaning of "extending the verification phase / reducing the verification phase speed" is: to expose the risks to a more moderate and controllable verification process;

[0168] After optimizing the convergence mechanism, the collaborative judgment result is updated. If the updated collaborative judgment result is still in the verification failure state more than 3 times, the central control system performs safety degradation processing. Safety degradation processing includes stopping the main transport section movement and retracting to a safe height, performing vacuum release and vacuum breaking actions to remove adsorption and prevent dragging, writing the abnormal working conditions, corresponding workstation numbers, and ambient temperature ranges into the data recording unit, and selecting manual intervention according to the upper computer strategy to perform re-grab and wait for defrosting, etc.

[0169] In this embodiment, method S41 first uses the historical collaborative risk amount Ram to extract the "acceptable upper limit of coupling risk" under the melting window condition as the collaborative risk threshold Rth. The purpose is to upgrade the release condition from "whether the single vacuum is sufficient" to "whether the comprehensive stability of vacuum fluctuation + starting impact meets the standard". Because in the water film stage, "the vacuum seems very stable but the friction is very low". If the starting impact is not included in the judgment, the robot may slip as soon as it enters the main transport section. Therefore, using Ram≤Rth as the release standard can screen out high-risk working conditions in the short-distance verification section and avoid material dropping and collision. If Ram > Rth triggers optimization, step S42 directly takes the absolute values ​​of the vacuum pressure variance and the terminal acceleration rate of change within W2, and uses "1.5 times threshold" to determine the dominant term. The purpose is to quickly pinpoint the root cause of the problem and prevent repeated switching when the two terms are close: for example, if the vacuum pressure variance is significantly larger, it indicates that the seal is shaking, and continuing to accelerate will only amplify leakage / micro-slippage. Therefore, 20ms PWM is used, the duty cycle is gradually reduced by 15% (60%→45%), and the stable holding time is set to 80ms. The lateral acceleration limit was gradually increased (120ms → 200ms) to allow for smoother vacuum establishment and stabilization time. When the initial impact was dominant, it indicated that the action was too forceful, pushing the water film interface to slip. Therefore, the upper limit of lateral acceleration was gradually decreased by 0.4 m / s² (1.2 → 0.8), the upper limit of acceleration change rate was gradually decreased by 3 m / s³ (8 → 5 → 2), and the length of the verification section was gradually increased by 30 mm (50 → 80 → 110) to expose the risk to a gentler, more controllable short distance. The direct effect of this setting is that, without changing the hardware, the control parameters converge interpretably "in the direction of reducing Ram," reducing blind repetitions after a failed verification. If Ram cannot be brought below the threshold after three consecutive attempts, a safety degradation process is implemented, reverting to a safe height and releasing the vacuum to avoid dragging and damage.

[0170] Example 6

[0171] Please see Figure 1 and Figure 2 An adjustment and control system for a suction cup fork-tooth loading and unloading robot includes an observation window establishment module, a frost layer melting analysis module, a collaborative analysis module, and an optimization convergence and degradation module.

[0172] The observation window establishment module sets up collection points and uses sensor groups set up at the collection points to collect operational data in real time and transmit it to the central control system, where the observation window is established.

[0173] The frost melting analysis module extracts features from the running data within the observation window to obtain window feature data. Based on the window feature data, it calculates the frost melting discriminant Cf by combining the sampling period Δt and the number of sampling points N1 in the window. Based on the output of the frost melting discriminant Cf, it conducts a preliminary comparative evaluation.

[0174] After the preliminary comparative evaluation determines that the frost layer is entering the melting window, the collaborative analysis module executes the adsorption and motion collaborative limiting control mechanism, and calculates the collaborative risk quantity Ram based on the frost layer melting discrimination quantity Cf within the observation window;

[0175] The optimization convergence and degradation module performs collaborative judgment based on a preset collaborative risk threshold Rth and collaborative risk amount Ram, and executes optimization convergence mechanism and security degradation based on the judgment result.

[0176] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for adjusting and controlling a suction cup fork-tooth loading and unloading robot, characterized in that: Includes the following steps: S1. By setting up collection points and using the sensor groups set up at the collection points, the operating data is collected in real time and transmitted to the central control system, where an observation window is established. S2. Extract features from the running data within the observation window to obtain window feature data. Calculate the frost melting discriminant Cf based on the window feature data, combined with the sampling period Δt and the number of sampling points N1. Perform a preliminary comparative evaluation based on the output of the frost melting discriminant Cf. The frost layer melting discrimination value Cf is calculated using the following algorithm formula: In the formula, Pv,start represents the average vacuum pressure of the first half of the window corresponding to the half-sampling point set, Pv,end represents the average vacuum pressure of the second half of the window corresponding to the second half of the sampling point set, Ts,start represents the average surface temperature of the first half of the window corresponding to the first half of the sampling point set, and Ts,end represents the average surface temperature of the second half of the window corresponding to the second half of the sampling point set. S3. After preliminary comparative evaluation and determination that the frost layer has entered the melting window, the adsorption and motion synergistic limiting control mechanism is implemented, and the synergistic risk quantity Ram is calculated based on the frost layer melting discrimination quantity Cf within the observation window. The collaborative risk quantity Ram is calculated and output using the following algorithm formula: In the formula, d represents the differential symbol, dt represents the differential with respect to the time variable, Var represents the variance, and Pv(t) represents the vacuum pressure at the suction end at time t. S4. The preset collaborative risk threshold Rth and collaborative risk quantity Ram are used to make a collaborative judgment, and the optimization convergence mechanism and security degradation are executed based on the judgment result.

2. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 1, characterized in that: S1 includes S11; S11. By setting several collection points on the suction cup fork loading and unloading robot and setting sensor groups in the collection points, the operation data of the suction cup fork loading and unloading robot during the adsorption and operation process is collected in real time, and the operation data is transmitted to the central control system of the suction cup fork loading and unloading robot through the wireless communication network. The data acquisition points include vacuum pressure acquisition points, surface temperature acquisition points, and motion state acquisition points; The operational data includes suction cup vacuum pressure sequence, workpiece surface temperature sequence, end-effector acceleration time sequence, and end-effector Z-axis position time sequence; Among them, each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence and end Z-direction position time sequence specifically includes the suction cup end vacuum pressure Pv(t) at time t, the workpiece adsorption area surface temperature Ts(t) at time t, the end acceleration a(t) at time t and the end Z-direction position z(t) at time t. The sensor group includes a vacuum pressure sensor, an infrared temperature sensor, and a robotic arm controller.

3. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 2, characterized in that: S1 further includes S12; S12. The central control system receives operational data in real time and performs data access processing on the operational data through the time synchronization unit of the central control system. The data access processing includes channel mapping of data from different sources, data integrity verification, and unified timestamp calibration. The time synchronization unit unifies the zero time with the adsorption establishment command trigger time and aligns the time axis of each time series in the operational data. Among them, when the sampling frequency of the surface temperature time of the workpiece adsorption region is lower than the sampling frequency of the vacuum pressure time, nearest neighbor alignment is used to align the surface temperature time series of the workpiece adsorption region to the sampling timestamp of the vacuum pressure time series. After the time axis alignment is completed, an adsorption establishment observation window W1 is established. The start time of the adsorption establishment observation window W1 is the trigger time of the adsorption establishment command. The duration of the adsorption establishment observation window W1 is determined by the preset number of sampling points N1 and the sampling period Δt. Then, for each sequence point in the suction cup vacuum pressure sequence, workpiece surface temperature sequence, end acceleration time sequence, and end Z-axis position time sequence, the suction cup end vacuum pressure Pv(t), workpiece adsorption area surface temperature Ts(t), end Z-axis position z(t), and end acceleration a(t) at time t are subjected to dimensionless processing to eliminate the influence of unit dimensions. The dimensionless processing is performed by normalizing the sampled values ​​based on the baseline mean and fluctuation range of each time sequence within the corresponding window to eliminate the influence of unit dimensions of each parameter.

4. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 3, characterized in that: S2 includes S21; S21. Based on the time series of vacuum pressure and the time series of surface temperature of the workpiece adsorption area collected within the adsorption observation window W1; within the adsorption observation window W1, the collection points are divided into a first half sampling point set and a second half sampling point set according to time sequence. The average vacuum pressure Pv,start of the first half sampling point set and the average vacuum pressure Pv,end of the second half sampling point set are calculated respectively. The average surface temperature Ts,start of the first half sampling point set and the average surface temperature Ts,end of the second half sampling point set are also calculated respectively. Based on the sampling period Δt and the number of sampling points N1 of the adsorption observation window W1, the frost layer melting discriminant Cf is calculated to quantitatively analyze the coupling relationship between the intensity of vacuum change and the surface temperature rise trend during the adsorption establishment stage.

5. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 4, characterized in that: S2 further includes S22; S22. Based on the output of the frost layer melting discrimination value Cf in the central control system, the central control system performs preliminary comparative evaluation processing to determine the transition stage of the frost layer from hard and brittle to melting; the specific evaluation content is as follows: When the frost layer melting discrimination value Cf>1, it indicates that the vacuum change is drastic and the temperature rise is insufficient. It is determined that the workpiece adsorption interface is in a state dominated by hard and brittle frost layer, and the evaluation result of not entering the frost layer melting window is output. The suction cup fork tooth loading and unloading robot is controlled to enter the micro-pulse adsorption and defrost mode and return to S2. When the frost layer melting discrimination value Cf≤1, it indicates that the temperature change begins to be significant. The workpiece adsorption interface is determined to have entered the frost layer melting window and the evaluation result of entering the frost layer melting window is output, thereby triggering the adsorption and motion coordinated limiting control mechanism.

6. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 5, characterized in that: S3 includes S31; S31. Under the condition of preliminary comparative evaluation of the adsorption and motion coordinated limiting control mechanism, the central control system sends an adsorption establishment slope shaping control command to the vacuum valve control unit. The adsorption establishment slope shaping control command shapes the slope of the vacuum establishment process during the adsorption establishment stage, so that the rise rate of the suction cup vacuum pressure sequence during the adsorption establishment period is limited and changes smoothly. The central control system sends a short-distance verification segment motion command to the robot controller, causing the suction cup fork-tooth loading and unloading robot to execute the short-distance verification segment motion after completing pre-contact micro-pressure and entering the adsorption establishment state. The short-distance verification segment motion applies coordinated amplitude limiting constraints to the motion planning parameters through the central control system. The coordinated amplitude limiting constraints include at least a lateral acceleration upper limit constraint and an acceleration change rate upper limit constraint. The upper limit constraint on lateral acceleration limits the maximum acceleration along the horizontal plane during short-distance verification segment motion. The upper limit constraint on the rate of change of acceleration limits the maximum rate of change of acceleration over time during the short-distance verification segment motion. Then, based on the starting moment of the short-distance verification segment motion, a verification segment evaluation window W2 is established, and the vacuum pressure time series and the terminal acceleration time series are collected within the verification segment evaluation window W2.

7. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 6, characterized in that: S3 further includes S32; S32. Establish a verification segment evaluation window W2 at the start moment of the short-distance verification segment motion, and collect the suction cup vacuum pressure sequence and the end acceleration time sequence within the verification segment evaluation window W2. Then, based on the suction cup vacuum pressure Pv(t) and the terminal Z-axis position z(t) at time t of the corresponding sequence point in the suction cup vacuum pressure sequence and the terminal acceleration time sequence, and after dimensionless processing, combined with the frost layer melting discrimination quantity Cf, the cooperative risk quantity Ram is calculated and output to quantitatively analyze the degree of coupling risk between vacuum fluctuation and start-up impact under the frost layer melting window condition.

8. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 7, characterized in that: S4 includes S41; S41. Based on the historical collaborative risk quantity Ram, extract the critical value of the coupling risk between vacuum fluctuation and initial impact under the frost melting window condition and set it as the collaborative risk threshold Rth. The real-time acquired collaborative risk quantity Ram is then used to make a collaborative judgment with the collaborative risk threshold Rth, and the collaborative judgment result is output. The specific judgment content is as follows: When the collaborative risk level Ram ≤ collaborative risk threshold Rth, the verification result is output and the suction cup fork loading and unloading robot is allowed to switch from the short-distance verification section movement to the main transport section movement. When the collaborative risk amount Ram is greater than the collaborative risk threshold Rth, the verification failure result is output and the optimization convergence mechanism is triggered.

9. The adjustment and control method for a suction cup fork-tooth loading and unloading robot according to claim 8, characterized in that: S4 further includes S42; S42. After triggering the optimization convergence mechanism, the central control system obtains the absolute values ​​of the vacuum pressure variance and the terminal acceleration change rate within the verification section evaluation window W2; and uses the dominant decision rule to determine the dominant term, and executes the corresponding optimization convergence mechanism based on the determined dominant term. The dominant determination rules are as follows: When the variance of vacuum pressure is ≥1.5 × the absolute value of the rate of change of terminal acceleration, it is determined to be the dominant term of vacuum fluctuation, and adsorption-side convergence is performed. When the absolute value of the rate of change of terminal acceleration is ≥1.5×vacuum pressure variance, it is determined to be the dominant term of the initial impact and motion-side convergence is performed. If none of the inequalities are satisfied, it is determined to be a coupling-dominant term, and both adsorption-side convergence and motion-side convergence are performed simultaneously. The adsorption-side convergence is achieved by the central control system adjusting the pulse width modulation slope shaping and stabilization holding time of the vacuum solenoid valve according to the gear position; the specific convergence process is as follows: Set the pulse width modulation period to 20ms; The effective duty cycle will be reduced by one level from the current level, with each level being 15%. Increase the adsorption stability time by one level from the current level, with each level being 80ms; The duty cycle is kept constant during the adsorption stabilization period to allow vacuum pressure fluctuations to converge. The motion-side convergence is achieved by the central control system performing coordinated amplitude-limiting convergence of motion parameters in the short-distance verification segment according to gear position. The specific convergence content is as follows: The upper limit of lateral acceleration is reduced by one level from the current level, with each level being 0.

4. The upper limit of the rate of change of acceleration is reduced by one level from the current level, with each level being 3 m / s². 3 ; Increase the length of the short-distance verification section by one level, with each level being 30mm. After optimizing the convergence mechanism, the collaborative judgment result is updated. If the updated collaborative judgment result is still in the judgment result of verification failure more than 3 times, the central control system performs safety degradation processing. The safety degradation processing includes stopping the movement of the main transport section and retracting to a safe height, performing vacuum release and vacuum breaking actions to remove adsorption and prevent dragging, writing the abnormal working conditions and corresponding workstation numbers and ambient temperature ranges into the data recording unit, and selecting manual intervention according to the upper computer strategy.

10. An adjustment and control system for a suction cup fork-tooth loading and unloading robot, applied to the adjustment and control method for a suction cup fork-tooth loading and unloading robot as described in any one of claims 1-9, characterized in that: It includes modules for establishing observation windows, analyzing frost melting, conducting collaborative analysis, and optimizing convergence and degradation. The observation window establishment module sets up collection points and uses sensor groups set up at the collection points to collect operational data in real time and transmit it to the central control system, where the observation window is established. The frost melting analysis module extracts features from the running data within the observation window to obtain window feature data. Based on the window feature data, it calculates the frost melting discriminant Cf by combining the sampling period Δt and the number of sampling points N1 in the window. Based on the output of the frost melting discriminant Cf, it performs a preliminary comparative evaluation. The collaborative analysis module, after determining the entry into the frost melting window through preliminary comparative evaluation, executes the adsorption and motion collaborative limiting control mechanism, and calculates the collaborative risk quantity Ram based on the frost melting discrimination quantity Cf within the observation window; The optimization convergence and degradation module performs a collaborative determination based on a preset collaborative risk threshold Rth and a collaborative risk amount Ram, and executes an optimization convergence mechanism and a security degradation based on the determination result.

Images