Intelligent control method and system for port bulk cargo handling equipment based on protocol data

Abstract

The application relates to the technical field of intelligent control of port hoisting and transporting equipment, and discloses a port bulk cargo handling equipment intelligent control method and system based on protocol data, which jointly establishes a practical residual mass benchmark of a grab bucket by means of unloading residual release and reverse paired acceleration verification, so that the bucket opening action and the load real release process are mutually checked; in the return phase, the mechanical state is further competed by utilizing the suspension mass platform, the residual error of the rope length and the torque residual error periodic characteristics, the delayed shedding is separated from the steel wire rope loss of tension and the load swing, and only the corresponding control is executed on the confirmed state; in view of the parameter pollution caused by the delayed shedding, the state conversion interval is isolated as a prohibited inheritance interval, the anti-swing parameter is identified only from the data with stable mass benchmark and supporting rope continuous tension, meanwhile, the empty bucket load benchmark is handed over to an independent empty bucket task for updating, so that the abnormal load data cross-task inheritance is blocked, and the same transient state is prevented from being repeatedly interpreted and written into the subsequent control model.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology for port lifting and transportation equipment, and in particular to intelligent control methods and systems for port bulk cargo loading and unloading equipment based on protocol data. Background Technology

[0002] In port operations, bulk materials such as grain and mineral powder are typically handled by grab bucket gantry cranes for retrieval from ship holds, aerial transfer, and unloading from hoppers. After the grab bucket moves above the hopper, the control system drives the opening and closing mechanism to open the jaws. Based on feedback from the opening mechanism, it determines whether the unloading phase is complete. Subsequently, it controls the lifting mechanism to raise the grab bucket, and then sends it back above the ship hold via the slewing and luffing mechanisms. When the material is dry and has good flowability, the material inside the grab bucket can fall into the hopper quickly under gravity after the jaws open. In actual port operations, the same batch of bulk material may experience rain, low-temperature storage, ship hold compaction, and repeated grabbing. Some material may adhere to the inner wall of the grab bucket, near the reinforcing ribs, or in the jaw plate rotation connection area under the squeezing action of the jaws. For example, during the unloading of water-containing mineral powder, when the grab bucket completes the opening action, most of the material has already fallen into the hopper, but locally compacted material may still remain attached and enter the lifting and slewing process with the grab bucket until it suddenly falls off after the grab bucket's posture, vibration state, or centrifugal force changes.

[0003] In a common control method, the control system switches the grab bucket from the heavy-load unloading stage to the empty bucket return stage based on the opening limit or the position of the opening and closing mechanism, and activates the lifting, slewing, luffing, and anti-sway control parameters under empty bucket conditions. This process can reflect that the grab bucket jaws have reached the open position. However, when material adheres, there may be a deviation between the completion time of the mechanism action and the actual load release time. The residual material suddenly falls off after the grab bucket enters the return stage, causing the support rope force and the grab bucket center of gravity to change synchronously in a short period of time. This response may form similar characteristics to the load swing or the change in wire rope tension. If the control system directly adjusts the lifting compensation, anti-sway parameters, or empty bucket mass model based on this transient response, the same load change may trigger different control actions at the same time, and the data of the delayed detachment stage will be included in the anti-sway parameters of the subsequent loading and unloading cycle. Therefore, it is necessary to further identify whether the grab bucket load is actually released after the bucket opening action is completed, and when the residual material falls off with a delay, determine the mechanical state corresponding to the sudden change in load and its application range to the subsequent motion control parameters. Summary of the Invention

[0004] This application proposes an intelligent control method and system for port bulk cargo handling equipment based on protocol data to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this application adopts the following technical solution: an intelligent control method for port bulk cargo handling equipment based on protocol data, comprising: Acquire grab bucket unloading protocol data, form unloading operation segments according to task identifiers and mechanical response intervals, perform transmission compensation for support rope drive torque, establish load release trajectory according to jaw plate opening, and generate unloading release residual. Based on the unloading release residual, the hoisting verification is initiated. When the slewing mechanism and the luffing mechanism are stationary and the hoisting direction remains unchanged, the hoisting mechanism executes the reverse paired acceleration segment. The paired torque response is processed by transmission conversion, length compensation and differential processing to generate the remaining suspended mass verification value, unloading confirmation status and return initial mass benchmark. Acquire return data, predict the normal support rope driving torque based on the initial return mass benchmark, extract the torque residual, reconstruct the suspended mass platform, and perform mechanical state competition by combining the changes in the suspended mass platform, the relationship between the drum rope length and the grab bucket displacement, and the periodic characteristics of the torque residual, generating mechanical states, including delayed detachment, wire rope loss of tension, load swaying, and unconfirmed state; in the case of delayed detachment, limit slewing and luffing and generate a mass benchmark after load change; in the case of wire rope loss of tension, stop rope release and perform tension protection; in the case of load swaying, generate slewing anti-sway correction and luffing anti-sway correction; Based on the mechanical condition, the system is divided into prohibited inheritance intervals and continuous tension intervals. Candidate anti-sway parameters are identified from the continuous tension intervals. The lateral displacement evaluation and support rope driving torque fluctuation evaluation are compared between the candidate anti-sway parameters and the benchmark anti-sway parameters under the same foundation suspension point acceleration and the same return state conditions. The benchmark anti-sway parameters are updated when both evaluation values ​​corresponding to the candidate anti-sway parameters are lower than those corresponding to the benchmark anti-sway parameters. The air bucket load benchmark is updated based on the fully open bucket load formed by multiple independently confirmed air bucket missions.

[0006] Furthermore, the specific operations for forming the unloading section are as follows: Based on the task identifier associated with the bucket opening control record and mechanism feedback record, the first moment when the position feedback of the opening and closing mechanism continuously changes along the bucket opening direction is determined as the mechanical response start point, and the moment when the jaw plate reaches the fully open bucket position and the equivalent load of the support rope meets the load stability condition within the continuous sampling interval is determined as the mechanical response end point.

[0007] Furthermore, the specific operation for generating the unloading release residual is as follows: Based on the hoisting transmission chain parameters and drum motion state, the support rope driving torque is converted into an equivalent load of the support rope, and the difference is made with the empty bucket opening reference load that matches the support rope length, luffing position, and bucket opening speed. The resulting additional load is grouped into load nodes according to the jaw plate opening interval. Under the constraint that the load nodes do not increase with the overall jaw plate opening, the load release trajectory is reconstructed. The unloading release residual is generated based on the continuous stable load record after the bucket is fully opened.

[0008] Furthermore, the specific procedures for performing the lifting verification are as follows: When the grab bucket is located in the safe verification zone of the bucket, the support rope is kept taut, and the slewing mechanism and the luffing mechanism are at zero speed, if the unloading release residual is not higher than the empty bucket residual boundary, the lifting verification is performed according to the verification acceleration corresponding to the empty bucket residual boundary; if the unloading release residual is not lower than the residual load boundary, the lifting verification is performed according to the verification acceleration corresponding to the residual load boundary; if the unloading release residual is located between the two boundaries, the verification acceleration is continuously configured according to the position of the unloading release residual between the two boundaries. When the lifting mechanism remains in the same lifting direction, it sequentially executes two acceleration segments with opposite acceleration directions and equal absolute values ​​of acceleration time integrals.

[0009] Furthermore, the specific operations for generating the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark are as follows: The torque responses of the lifting speed difference, support rope length difference and grab bucket swing angle difference in the two acceleration segments, which are respectively within the corresponding pairing allowable range, are reverse-paired. The average support rope length of each pairing sample is used as the pairing reference length. The paired torque responses are converted into load responses under the pairing reference length. The dynamic suspension mass is identified based on the obtained paired load responses and paired actual acceleration responses. When the absolute value of the difference between the stable load before and after verification does not exceed the platform drift allowance, the dynamic suspended mass is subtracted from the dynamic mass benchmark of the empty bucket under the corresponding working condition; when the absolute value of the obtained mass difference does not exceed the allowable boundary of the empty bucket mass, the remaining suspended mass verification value is determined to be zero; when the mass difference is higher than the allowable boundary of the empty bucket mass, the mass difference is determined to be the remaining suspended mass verification value; when the mass difference is lower than the negative number of the allowable boundary of the empty bucket mass, the unloading confirmation state is determined to be unconfirmable; when the absolute value of the difference between the stable load before and after verification exceeds the platform drift allowance, the unloading confirmation state is determined to be unconfirmable. When the unloading status is confirmed to allow the return trip, the sum of the unloaded mass of the grab bucket and the verification value of the remaining suspended mass is determined as the initial mass benchmark for the return trip.

[0010] Furthermore, the specific operations for extracting the torque residuals and reconstructing the suspended mass platform are as follows: Based on the actual rope winding structure of the support rope, the amount of rope released from the drum, the length of the tensioned rope path from the grab bucket to the suspension point of the support rope, and the elastic elongation of the support rope are converted to the same support rope ratio to form the rope length bearing residual. Based on the initial mass benchmark of the return trip, the dynamic equivalent mass of the wire rope that varies with the length of the support rope and the hoisting acceleration, and the acceleration of the grab bucket along the support rope direction, the equivalent load of the normal support rope is predicted. The equivalent load of the normal support rope is converted into the driving torque of the normal support rope according to the hoisting transmission chain relationship. The difference between the measured driving torque of the support rope and the driving torque of the normal support rope is calculated to generate the torque residual. When the support rope is kept taut and the rope length bearing residual is within the tension tolerance zone, the instantaneous suspended mass is formed based on the measured equivalent load of the support rope, the acceleration of the grab bucket along the support rope direction, and the dynamic equivalent mass of the wire rope. The suspended mass platform is then formed by recursively using the current and historical instantaneous suspended masses.

[0011] Furthermore, the specific operations for executing mechanical state competition are as follows: Before the delayed detachment is confirmed, the initial quality benchmark of the return trip is determined as the current quality benchmark; after the delayed detachment is confirmed, the quality benchmark after the most recent load change is determined as the current quality benchmark. When the equivalent load of the support rope enters the untension protection range and the rope length bearing residual continuously exceeds the tension allowable zone, the wire rope is confirmed to be untensioned, the drum is stopped from continuing to release the rope and tension protection is implemented. If the wire rope is not confirmed to be out of tension, and the suspended mass platform does not recover to the current mass benchmark within one swing cycle after descent, the rope length bearing residual is within the tension tolerance zone, and the opening and closing mechanism does not generate new bucket opening and closing actions, then the delayed detachment is confirmed. The stable suspended mass platform after descent is determined as the mass benchmark after the load change, and the acceleration of the slewing mechanism and the luffing mechanism is limited. If the wire rope is not confirmed to be out of tension, and the suspended mass platform does not form a step change that continuously deviates from the current mass benchmark within one swing cycle, the torque residual has swing cycle characteristics, and the lateral velocity of the grab bucket reverses direction, the load swing is confirmed, and the slewing anti-sway correction amount and the amplitude anti-sway correction amount are generated according to the lateral displacement and lateral velocity of the grab bucket. When delayed detachment, wire rope slack, and load oscillation cannot be mutually exclusively confirmed, the mechanical state is determined as unconfirmable, the restricted control of the slewing mechanism and luffing mechanism is maintained, and the current quality benchmark is stopped from being updated.

[0012] Furthermore, the specific steps for dividing the prohibited inheritance interval and the continuous tension interval are as follows: The delayed detachment quality conversion range, wire rope loss and retensioning range, load swing entry and exit range, unconfirmed state range and actuator interlock range are divided into prohibited inheritance ranges. The boundaries of the prohibited inheritance ranges are expanded according to the maximum data age of the data used to determine the corresponding mechanical state and the state recovery stabilization time. After deducting the prohibited inheritance interval from the return data, extract the continuous tension interval where the support rope is continuously tensioned, the rope length bearing residual is within the tension tolerance zone, the current quality benchmark remains unchanged, the protocol data is continuous, and the anti-sway control mode remains unchanged. Then, group the intervals according to the current quality benchmark, support rope length working condition, and anti-sway control mode.

[0013] Furthermore, the specific steps for updating the baseline anti-sway parameters and the empty bucket load baseline are as follows: From the continuous tension intervals in the same group, select non-overlapping parameter identification intervals and parameter verification intervals, and identify candidate anti-sway parameters under the condition that the anti-sway model structure, equipment structure parameters and current quality benchmark remain unchanged. Candidate anti-sway parameters and benchmark anti-sway parameters are used respectively. Under the same basic suspension point acceleration, initial grab state, support rope length sequence, current mass benchmark, and external disturbance sequence, a closed-loop prediction model is run to generate lateral displacement evaluation quantity and support rope driving torque fluctuation evaluation quantity. When both evaluation quantities corresponding to the candidate anti-sway parameters are lower than the evaluation quantities corresponding to the benchmark anti-sway parameters, and the suspended mass platform is consistent with the current mass benchmark, the benchmark anti-sway parameters and candidate anti-sway parameters are weighted and synthesized according to the update ratio determined by the upper limit of the allowable single benchmark parameter change and the maximum allowable deviation of the candidate anti-sway parameters from the benchmark anti-sway parameters, and the benchmark anti-sway parameters are updated. Each independently confirmed air bucket mission generates a candidate value for fully open bucket load; after generating multiple candidate values ​​for fully open bucket load under the same working condition, the median value of the candidate values ​​is calculated. When the absolute difference between each candidate value and the median value does not exceed the load consistency boundary, the air bucket load benchmark for the corresponding working condition is updated.

[0014] The intelligent control system for port bulk cargo handling equipment based on protocol data includes: The protocol data acquisition module communicates with the hoisting drive, opening and closing drive, drum encoder, slewing drive, luffing drive and grab bucket position detection device to acquire grab bucket unloading protocol data and return data, and group them according to task identifier and unified control time scale. The unloading process analysis module forms the unloading operation section, performs transmission compensation on the drive torque of the support rope, establishes the load release trajectory according to the jaw plate opening, and generates the unloading release residual. The lifting verification module verifies the acceleration based on the unloading release residual configuration, controls the lifting mechanism to execute the reverse paired acceleration segment, and generates the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark. The return state recognition module generates rope length bearing residuals and torque residuals, reconstructs the suspended mass platform, executes mechanical state competition, and controls the hoisting mechanism, slewing mechanism, and luffing mechanism according to the mechanical state. The parameter inheritance module divides the prohibited inheritance interval and the continuous tension interval, identifies and verifies the candidate anti-sway parameters, and updates the benchmark anti-sway parameters and the empty bucket load benchmark respectively.

[0015] The beneficial effects of this invention are as follows: This invention establishes the actual remaining mass benchmark of the grab bucket by jointly verifying the unloading release residual and the reverse paired acceleration, enabling mutual verification between the bucket opening action and the actual load release process. During the return phase, it further utilizes the periodic characteristics of the suspended mass platform, rope length bearing residual, and torque residual to conduct mechanical state competition, separating delayed detachment from wire rope loss and load sway, and only performing corresponding control on confirmed states. To address parameter contamination caused by delayed detachment, this invention isolates the state transition interval into a prohibited inheritance interval, identifying anti-sway parameters only from data where the mass benchmark is stable and supports continuous rope tension. At the same time, the empty bucket load benchmark is handed over to an independently confirmed empty bucket task for updating, thereby blocking the cross-task inheritance of abnormal load data and avoiding the same transient being repeatedly interpreted and written into the subsequent control model. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort: Figure 1 This is a schematic diagram illustrating the overall principle of the present invention; Figure 2 This is a schematic diagram illustrating the construction of the unloading operation section and load release trajectory of the present invention; Figure 3 This is a schematic diagram illustrating the reverse paired acceleration lifting verification principle of the present invention. Detailed Implementation

[0017] 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.

[0018] like Figure 1 , Figure 2 and Figure 3 As shown, this invention discloses an intelligent control method for port bulk cargo handling equipment based on protocol data, including the following specific steps: This embodiment is applied to a four-rope grab bucket gantry crane at a bulk grain terminal. The gantry crane drives the grab bucket through the support rope and the opening and closing rope to unload the bulk grain material in the hold into a fixed bucket. When the bulk grain material clumps locally, some material may remain inside the jaw plate and between the bucket teeth after the jaw plate reaches the fully open position, and may fall off late during the grab bucket's return journey.

[0019] The gantry crane is equipped with a multi-axis automatic controller, a programmable controller, a support rope hoisting driver, an opening and closing driver, a drum encoder, a slewing driver, a luffing driver, an automatic tracking camera for the grab bucket, and a 3D point cloud monitoring radar. The programmable controller acquires the position of the opening and closing mechanism, the support rope driving torque, the drum rotation angle, the hoisting speed, the slewing position, the luffing position, and the operating status of the mechanism; the automatic tracking camera for the grab bucket and the 3D point cloud monitoring radar acquire the spatial position of the grab bucket.

[0020] This embodiment takes a gantry crane with a single support rope drive channel driving a support rope drum as an example. When the equipment is equipped with multiple independent support rope drive channels, torque compensation and rope tension conversion are performed on each drive channel respectively. Then, the normal drive torque of each channel is calculated based on the channel load distribution ratio formed by the initial empty bucket calibration. The torque of multiple drive channels is not directly calculated by using a total load value.

[0021] The current unloading task begins when the bucket opening command is authorized and ends when the task identifier changes; the current unloading task also ends when the task identifier remains unchanged and the next bucket closing command is authorized. Within the current unloading task, the unloading operation section, the hoisting verification section, and the return operation section are formed in sequence.

[0022] The protocol records carry the task identifier, device number, source timestamp, control cycle number, and validity status. Records whose task identifier, device number, and control cycle number are all consistent with the current task are included in the current calculation. When there are multiple records with the same source in the same control cycle, the record with the latest source timestamp and a normal validity status is retained. Records with backward source timestamps, duplicate control cycle numbers, or abnormal validity status are not included in the current calculation.

[0023] Each data source converts its source timestamp to a unified control timestamp through the controller clock synchronization relationship. The difference between the controller's received time and the converted source timestamp is used as the communication transmission delay. The communication transmission delay is used to calculate the data age, without changing the mechanical occurrence time corresponding to the feedback data. When the data age exceeds twice the update cycle of the corresponding data source, the data is marked as invalid.

[0024] During the debugging phase, the bucket opening command, hoisting acceleration command, slewing command, and luffing command are triggered repeatedly at least 10 times. The time difference between the moment the command is issued and the corresponding mechanism's response changes along the command direction for three consecutive sampling cycles is recorded. The median value of the effective time difference is taken as the command response delay of the corresponding mechanism. The command response delay is used to set the mechanical response search interval and control phase boundary, and is not used to modify the source timestamp of the feedback record. After deleting calibration records that do not meet the operating conditions or have discontinuous communication, if there are fewer than 5 effective calibration records, the channel will not participate in cross-channel data pairing.

[0025] Online control only uses valid data that has arrived before the current unified control time. It requires mechanical status confirmation and parameter verification for the entire time window. After the corresponding time window ends, it uses historical data within the window for execution. The confirmation result only applies to the control process after the confirmation time.

[0026] The programmable controller scanning cycle is read from the controller task configuration table, preferably 5ms to 20ms, and 10ms is used in this embodiment; the feedback cycle for the support rope drive torque and drum rotation angle is read from the driver communication configuration, preferably 10ms to 20ms, and 10ms is used in this embodiment; the position feedback cycle for the opening and closing mechanism is preferably 20ms to 50ms, and 20ms is used in this embodiment; the spatial position feedback cycle of the grab bucket is determined according to the effective output cycle of the grab bucket position detection device, preferably 40ms to 100ms, and 50ms is used in this embodiment. When the equipment configuration cycle changes, the number of samples in the window is reset according to the actual configuration cycle.

[0027] After the equipment is put into operation and maintenance involving the grab bucket, support rope, drum, and hoisting transmission chain is completed, the grab bucket is emptied while under maintenance. The grab bucket is kept fully open, no material outline is detected in the grab bucket image, and the volume occupied inside the bucket by the 3D point cloud is not higher than the minimum resolvable volume of the point cloud detection device. At this time, the initial empty bucket calibration task is registered. No less than 10 initial empty bucket calibration tasks are performed in the same target working condition unit. These tasks are used to form the empty bucket opening reference load, the empty bucket dynamic mass benchmark, the wire rope dynamic equivalent mass calibration node, the platform drift tolerance, and the load consistency boundary. The confirmed empty bucket tasks formed during the operation phase are used to update the existing benchmarks and do not constitute the initial initialization.

[0028] In this embodiment, the empty bucket load reference represents the equivalent load of the empty bucket support rope when the grab bucket is fully open under the same working condition; the empty bucket dynamic mass reference represents the dynamic suspended mass identified when the empty bucket grab bucket performs the reverse paired acceleration segment; the return trip initial mass reference represents the sum of the grab bucket empty mass and the remaining suspended mass verification value formed by the lifting verification; the current mass reference represents the mass of the grab bucket and the material inside the bucket used for normal load prediction in the current return trip stage, excluding the dynamic equivalent mass of the wire rope.

[0029] The following describes the continuous processing from steps S1 to S4 for a single bulk grain unloading task corresponding to task identifier T0427. The unloaded mass of the grab bucket is taken as 7800 kg, the effective length of the support rope is taken as 24.2 m, the luffing position is taken as 27.6 m, the lifting verification clearance height above the bucket is taken as 1.50 m, and the normal return lifting speed is taken as 0.35 m / s. The unloaded mass of the grab bucket is determined based on the equipment nameplate and the weighing record of the lifting device; the effective length of the support rope is calculated based on the drum position, the rope winding relationship, and the spatial position of the grab bucket; the luffing position is obtained by the luffing position sensor; the lifting verification clearance height is calculated based on the bucket installation dimensions and the current position of the grab bucket. These values ​​are a set of preferred calculated values ​​and do not represent the conclusions of actual field measurements.

[0030] S1 acquires the grab bucket unloading protocol data, forms unloading operation segments according to task identifiers and mechanical response intervals, performs transmission compensation on the support rope drive torque, establishes the load release trajectory according to the jaw plate opening, and generates unloading release residuals.

[0031] After obtaining permission to execute the bucket opening command, the position of the opening and closing mechanism, the driving torque of the support rope, the drum rotation angle, and the state of the hoisting mechanism are extracted from the current task. The current bucket opening stroke is obtained by subtracting the stable position before bucket opening from the current opening and closing mechanism position. The effective bucket opening stroke is obtained by subtracting the stable position before bucket opening from the fully bucket opening calibration position. The normalized jaw plate opening is obtained by dividing the current bucket opening stroke by the effective bucket opening stroke. The normalized jaw plate opening is set to 0 when it is less than 0 and 1 when it is greater than 1.

[0032] The mechanical response start point is determined by the opening start threshold, which is preferably 0.2% to 0.8% of the full stroke of the opening and closing mechanism, and not less than twice the position feedback resolution and the mechanism backlash calculation amount.

[0033] In a preferred implementation, the full stroke of the opening and closing mechanism is 1200mm, the position feedback resolution is 1mm, the mechanism backlash is 4mm, and 0.5% of the full stroke is 6mm, which is higher than twice the position feedback resolution and the mechanism backlash. Therefore, the opening start threshold is 6mm. When the position feedback of the opening and closing mechanism exceeds 6mm and continues to change along the opening direction in the subsequent two position feedback cycles, the moment when it first exceeds 6mm is determined as the mechanical response start point. The two consecutive position feedback cycles are used to eliminate the jitter of a single sampling point.

[0034] The allowable range for fully open bucket position is determined based on the repeatability error of the fully open bucket position and the backlash of the mechanism. It is preferably 0.3% to 1.0% of the full stroke of the opening and closing mechanism. In this embodiment, it is 6mm. When the absolute difference between the jaw plate position and the fully open bucket calibrated position does not exceed 6mm, the jaw plate enters the fully open bucket position range.

[0035] The support rope equivalent load stabilization window covers at least 5 support rope drive torque feedback cycles to form no less than 5 adjacent load changes. The upper limit of the window is determined based on the time required for the load to enter a stable state after the jaws are fully opened in the initial empty bucket calibration task. In this embodiment, the feedback cycle is 10ms, and the load stabilization time after the full opening of the bucket in the initial empty bucket calibration does not exceed 100ms. Therefore, the stabilization window is set to 100ms.

[0036] The load stability boundary is formed using adjacent load change samples from the fully open stable range in no less than 10 initial air bucket calibration missions. When the total number of effective change samples is no less than 100, the 95th percentile of the absolute value of the adjacent load change is taken, and then rounded up to an integer multiple of the support rope equivalent load resolution. In this embodiment, 0.8kN is used.

[0037] When the jaw plate remains within the fully open bucket position, the hoisting mechanism remains unchanged, and the equivalent load change of adjacent support ropes within a 100ms stabilization window does not exceed 0.8kN, the end time of the stabilization window is determined as the mechanical response endpoint; the time interval between the mechanical response start point and the mechanical response endpoint is determined as the mechanical response interval, and the records within the mechanical response interval form the unloading operation segment.

[0038] Transmission compensation is performed on the drive torque of the support rope in the unloading section. The output torque of the support rope hoisting motor is multiplied by the hoisting transmission ratio and transmission efficiency to obtain the drive torque on the drum side. The difference in angular velocity between adjacent drums is divided by the sampling time interval to obtain the drum angular acceleration. The drum angular acceleration is multiplied by the equivalent moment of inertia to obtain the inertia compensation torque. The Coulomb friction torque is read according to the drum rotation direction. The drum angular velocity is multiplied by the viscous friction coefficient to obtain the viscous friction torque.

[0039] The effective drum torque is obtained by subtracting the inertial compensation torque, Coulomb friction torque, and viscous friction torque from the drum-side driving torque. The effective drum radius is determined based on the nominal drum radius, wire rope diameter, and current number of winding layers. The effective drum torque is divided by the current effective drum radius to obtain the drum tangential force.

[0040] Based on the equipment's rope winding diagram, determine the supporting rope segments involved in the grab's load-bearing capacity. Add the projected components of the tension of each supporting rope segment relative to the lifting direction to obtain the grab's load-bearing capacity. The ratio of the grab's load-bearing capacity to the drum's tangential force is used as a dimensionless rope winding load conversion factor. Multiply the drum's tangential force by the rope winding load conversion factor to obtain the equivalent load of the supporting rope.

[0041] The transmission efficiency is calibrated through the constant speed lifting and constant speed lowering processes in the initial empty bucket calibration task, preferably between 0.88 and 0.96. In this embodiment, 0.92 is used. The equivalent moment of inertia, Coulomb friction torque, and viscous friction coefficient are calibrated through the low acceleration empty bucket lifting process to minimize the sum of the square errors between the compensated support rope equivalent load and the load corresponding to the empty mass of the grab bucket.

[0042] The reference working condition units for empty buckets are divided according to the length of the support rope, the position of the luffing change, and the opening speed. The grouping width of the support rope length is determined according to the fact that the change in the dynamic equivalent mass of the wire rope within the same group does not exceed the minimum resolution of the dynamic mass, preferably 1m to 3m, and 2m is used in this embodiment. The grouping width of the luffing change position is determined according to the fact that the change in the empty bucket load caused by the change in the suspension point of the support rope within the same group does not exceed 1 times the resolution of the equivalent load of the support rope, preferably 1.5m to 3m, and 2m is used in this embodiment. The opening speed is divided into low speed, normal speed and high speed according to the rated speed of the opening and closing mechanism.

[0043] If there are fewer than 5 valid initial empty bucket calibration tasks in the current working unit, the empty bucket opening reference load for that working unit will not be generated, data from other working units will not be called to replace it, and the current unloading confirmation status will be set to unconfirmed.

[0044] The number of jaw plate opening intervals should ensure that each opening interval still includes no less than 2 effective position feedbacks at the fastest bucket opening speed. The position feedback cycle of the opening and closing mechanism is 20ms. When no less than 48 effective position samples can be obtained during the complete bucket opening process, the normalized jaw plate opening from 0 to 1 is divided into 24 continuous opening intervals.

[0045] For the same initial air bucket calibration task, the median value of the equivalent load of the support rope falling into the same opening range is taken. Then, the median value of the corresponding median value of different initial air bucket calibration tasks in the same working condition unit is taken to obtain the air bucket reference node. When the current jaw plate opening is between two adjacent opening nodes, linear interpolation is performed according to the ratio of the jaw plate opening to the distance between the two nodes.

[0046] The empty bucket reference node in the fully open bucket range is determined as the empty bucket load reference. The empty bucket reference node in other opening ranges is subtracted from the empty bucket load reference to obtain the opening shape difference. The empty bucket opening reference load corresponding to the current jaw plate opening is equal to the sum of the empty bucket load reference and the corresponding opening shape difference.

[0047] For the effective sampling time in the unloading operation section, the equivalent load of the support rope is subtracted from the reference load of the empty bucket under the same jaw plate opening to obtain the additional load. The additional load is assigned to 24 opening intervals, and the median value of the additional load in each interval is taken to form the load node.

[0048] Starting from the beginning of the opening and moving towards the fully opened end, adjacent load nodes are compared sequentially. When the later load node is higher than the previous load node, the two opening intervals are merged, and the median value of all additional loads in the merged interval is used as the merged node value. If the merged node is still higher than the previous node, the merging continues. After all nodes stop increasing along the direction of jaw plate opening, the merged node value is assigned to the opening interval it covers, forming a load release trajectory.

[0049] After the jaw plate enters the fully open bucket position range, a terminal stability set is formed from a continuous record of the jaw plate opening remaining unchanged, the hoisting mechanism remaining unchanged, the support rope remaining taut, and the additional load meeting the stability conditions.

[0050] The terminal stabilization set includes at least 10 support rope drive torque feedback samples, so that the resulting load median value is not determined by a single instantaneous impact. Under a 10ms feedback cycle, the corresponding shortest continuous time is 100ms. The longest waiting time is rounded up based on the maximum time from full opening of the bucket to load stabilization in the initial empty bucket calibration task, and is limited by the allowable dwell time above the hopper. In this embodiment, it is taken as 2.0s.

[0051] If no 10 valid samples are obtained during the first waiting period, a full bucket opening and stable sampling will be performed again. If the number of samples still does not meet the requirements after the second sampling, the unloading confirmation status will be set to unconfirmed and the lifting verification will not proceed.

[0052] When at least 10 valid samples are obtained, the median value of the additional load in the terminal stable set is taken. If the absolute difference between the median value and the fully open terminal node of the load release trajectory does not exceed the load stability boundary, the median value of the terminal stable set is determined as the unloading release residual. If the absolute difference exceeds the load stability boundary, the current unloading operation section is marked as inconsistent data and the unloading release residual is not output.

[0053] In the preferred calculation example of task T0427, the equivalent load of the fully opened bucket support rope is 84.1kN. The current support rope length is in the 23m to 25m working unit, the amplitude change position is in the 27m to 29m working unit, the bucket opening speed is constant, and the corresponding empty bucket load benchmark is 79.0kN. The median value of the additional load in the terminal stability set is 5.1kN, and 5.1kN is determined as the unloading release residual.

[0054] S2 initiates the lifting verification based on the unloading release residual, controls the lifting mechanism to execute the reverse paired acceleration segment when the slewing mechanism and luffing mechanism are stationary and the lifting direction remains unchanged, calculates the difference in torque response, and generates the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark.

[0055] Lifting verification is permitted when the minimum distance between the grab bucket's outer contour and the boundary of the bucket's verification safety zone is not less than the safety distance, the support rope is kept taut, the lifting brake is released, and the slewing mechanism and luffing mechanism are in a zero-speed state.

[0056] The safety distance is formed by adding the grab position detection error, the maximum horizontal displacement during the verification period, and the equipment safety margin. The equipment safety margin is preferably 0.10m to 0.30m, and in this embodiment, it is 0.20m.

[0057] The zero-speed boundary of the slewing mechanism is taken as the larger of three times the slewing speed feedback resolution and the maximum speed fluctuation value in the braking holding state; the zero-speed boundary of the luffing mechanism is determined in the same way. In this embodiment, the slewing speed feedback resolution is 0.01° / s and the luffing speed feedback resolution is 0.003m / s. Therefore, when the absolute value of the slewing speed is not higher than 0.03° / s and the absolute value of the luffing speed is not higher than 0.01m / s, the slewing mechanism and the luffing mechanism are determined to be in a zero-speed state.

[0058] Divide the horizontal offset between the grab bucket and the suspension point of the support rope by the effective length of the support rope, and then take the arcsine of the resulting ratio to obtain the grab bucket swing angle. When the grab bucket swing angle is not higher than 2.0°, the verification is carried out. When the swing angle is 2.0°, the difference between the gravity projection coefficient and 1 is less than 0.001, and the converted mass error is lower than the allowable boundary of the empty bucket mass. Therefore, in this embodiment, the allowable value of the swing angle is taken as 2.0°.

[0059] The empty bucket residual boundary is formed by the unloading release residual of no less than 10 initial empty bucket calibration tasks. The maximum value of each unloading release residual is taken and superimposed with the upper limit of the transmission compensation residual to obtain the empty bucket residual boundary. In this embodiment, 2.8kN is taken.

[0060] The allowable limit of the dynamic mass of the air bucket is taken as 180 kg. The upper limit of the residual mass of the normal return trip is taken as 1.3 to 1.5 times the allowable limit of the dynamic mass of the air bucket, and then rounded up to an integer multiple of 50 kg. In this embodiment, it is taken as about 1.39 times 180 kg, resulting in 250 kg.

[0061] Multiplying 250kg by the gravitational acceleration yields approximately 2.45kN. Adding 2.45kN to the 2.8kN empty bucket residual boundary and rounding up to 0.1kN, we obtain a 5.3kN residual load boundary. The lifting verification is only performed when the grab bucket swing angle is no higher than 2.0°. The gravity projection difference caused by the swing angle is lower than the equivalent load resolution of the support rope. Therefore, the boundary is calculated based on 250kg multiplied by the gravitational acceleration.

[0062] The high-resolution verification acceleration is determined based on the pairwise response uncertainty of the minimum resolution of the bucket's dynamic mass and the equivalent load of the support rope, ensuring that the pairwise load response corresponding to the minimum resolution of the dynamic mass is not less than twice the response uncertainty, while guaranteeing that the lifting velocity does not change direction within the negative acceleration range. The low-impact verification acceleration is determined based on the minimum stable acceleration of the lifting actuator and the allowable inertial load of the residual material. In this embodiment, a high-resolution verification acceleration of 0.40 m / s² is used. 2 The low impact verification acceleration was 0.20 m / s². 2 .

[0063] When the unloading release residual is no higher than 2.8 kN, the verification acceleration is taken as 0.40 m / s². 2 When the unloading release residual is not less than 5.3 kN, the verification acceleration is taken as 0.20 m / s². 2 When the unloading release residual is between the two boundaries, subtract 2.8kN from the unloading release residual and divide by 2.5kN to obtain the low impact acceleration weight. Subtract the weight from 1 to obtain the high resolution acceleration weight. Multiply the two verification accelerations by their corresponding weights and add them together to obtain the current verification acceleration.

[0064] The unloading release residual for task T0427 is 5.1 kN, corresponding to a low-impact acceleration weight of 0.92, a high-resolution acceleration weight of 0.08, and a verification acceleration of 0.216 m / s². 2 In this embodiment, 0.22 m / s is used. 2 .

[0065] The duration of a single acceleration segment should allow at least 40 original samples to be obtained under a 10ms support rope drive torque feedback cycle, and the complete verification time should not exceed 1 / 8 of the theoretical swing cycle. In this embodiment, a single acceleration segment is taken as 0.45s, corresponding to 45 original samples. The velocity holding time between two acceleration segments should cover at least 5 support rope drive torque feedback cycles, which is taken as 0.10s in this embodiment.

[0066] The hoisting mechanism moves upward at a reference speed of 0.35 m / s, sequentially executing a positive acceleration segment lasting 0.45 s, a speed holding segment lasting 0.10 s, and a negative acceleration segment lasting 0.45 s. The current time within the acceleration segment is divided by 0.45 s to obtain the normalized process. The normalized process is multiplied by π and the sine value is calculated to obtain the acceleration shape coefficient. The positive acceleration segment command is to verify the product of acceleration and acceleration shape coefficient, and the negative acceleration segment command is the negative of the product.

[0067] The absolute values ​​of the acceleration time integrals in the two acceleration segments are equal. The negative acceleration segment only reduces the upward velocity, and the lifting direction remains unchanged during the verification process.

[0068] After both acceleration segments have ended, the dynamic suspension mass is calculated. For each sampling moment, the current lifting speed and the previous four lifting speed samples are used to perform linear fitting. The slope of the fitted line is taken as the actual vertical acceleration at the current moment, and data after the current moment are not used.

[0069] Samples in the positive acceleration segment with a normalized process value equal to the first process value are used to establish candidate pairings with samples in the negative acceleration segment with a normalized process value equal to 1 minus the first process value. The allowable range for rope length pairing is determined based on the effective interpolation range of the theoretical maximum pairing displacement formed by verifying the motion trajectory and the dynamic equivalent mass calibration relationship of the wire rope. This implementation uses a speed of 0.20 m / s². 2 up to 0.40m / s 2 The theoretical maximum paired displacement under acceleration was verified to be no higher than 0.42m. The initial empty bucket calibration confirmed that the dynamic equivalent mass of the wire rope could be effectively interpolated when the adjacent lengths did not exceed 0.50m. Therefore, the allowable range for the paired support rope length was taken as 0.45m; the allowable range for the paired lifting speed was taken as 0.03m / s; and the allowable range for the paired grab bucket swing angle was taken as 0.5°.

[0070] If the difference in lifting speed and the difference in grab bucket swing angle are within the allowable range of the corresponding pairing, the difference in support rope length does not exceed 0.45m, and the support rope lengths of the two samples are both within the coverage of the existing wire rope dynamic equivalent quality calibration node, then the candidate pairing is determined as a valid pairing sample.

[0071] For each valid paired sample, the average length of the support rope between the positive acceleration segment sample and the negative acceleration segment sample is taken as the paired reference length. Based on the dynamic equivalent mass calibration relationship of the wire rope, the dynamic equivalent mass of the wire rope corresponding to the acceleration level of the positive acceleration segment is obtained at both the actual support rope length and the paired reference length. The dynamic equivalent mass of the wire rope corresponding to the paired reference length is subtracted from the dynamic equivalent mass of the wire rope corresponding to the actual support rope length. The resulting mass difference is multiplied by the mass load of the positive acceleration segment sample to convert to acceleration, and then added to the equivalent load of the support rope in the positive acceleration segment to obtain the positive length compensation load.

[0072] Following the same method, based on the actual support rope length and paired reference length of the negative acceleration segment sample, and the dynamic equivalent mass of the wire rope corresponding to the acceleration level of the negative acceleration segment, the equivalent load of the support rope in the negative acceleration segment is converted to obtain the negative length compensation load. The mass load converted acceleration is formed by adding the product of gravitational acceleration and the cosine of the corresponding grab bucket swing angle, and the product of the actual vertical acceleration and the cosine of the corresponding grab bucket swing angle.

[0073] Subtracting the negative length compensation load from the positive length compensation load and dividing by 2 yields the paired load response. Multiplying the actual vertical acceleration of the positive acceleration segment by the cosine of the corresponding grab angle yields the positive acceleration response; multiplying the actual vertical acceleration of the negative acceleration segment by the cosine of the corresponding grab angle yields the negative acceleration response; subtracting the negative acceleration response from the positive acceleration response and dividing by 2 yields the paired actual acceleration response. Both the positive and negative acceleration responses retain the directional sign of the actual vertical acceleration, but the absolute value of the actual vertical acceleration in the negative acceleration segment is not taken.

[0074] Calculate the ratio of the difference in lifting speed to the allowable range of lifting speed pairing, the ratio of the difference in support rope length to the allowable range of support rope length pairing, and the ratio of the difference in grab bucket swing angle to the allowable range of grab bucket swing angle pairing for each pair of samples. Take the maximum value of the three ratios and subtract the maximum value from 1 to obtain the pairing weight; take 0 when the pairing weight is less than 0.

[0075] The first accumulated value is obtained by multiplying the paired load response, paired actual acceleration response and paired weight of each effective paired sample; the second accumulated value is obtained by multiplying the square of each paired actual acceleration response with the paired weight; and the dynamic suspension mass is obtained by dividing the first accumulated value by the second accumulated value.

[0076] Valid paired samples should cover the first, middle and last segments of the acceleration curve, with no less than 4 pairs in each segment. Therefore, the total number of valid paired samples should not be less than 12 pairs. The sum of squares of the actual accelerations of the pairs corresponding to the valid samples is divided by the sum of squares of the actual accelerations of the pairs of all available paired samples to obtain the acceleration energy coverage. The acceleration energy coverage should not be less than 85%, and the dynamic suspension mass is valid when the second accumulated value is higher than the 5th percentile of the second accumulated value in the initial air bucket calibration.

[0077] If any valid verification condition is not met, the full verification will be re-executed; if it is not met twice in a row, the unloading confirmation status will be set to unconfirmed and no return permit will be generated.

[0078] Before and after the verification, the equivalent load of the stable support rope is extracted for 100ms. In no less than 10 initial air bucket verifications, the absolute value of the difference between the stable load before and after the verification is calculated. The maximum value is taken and superimposed with 1 times the resolution of the support rope equivalent load to obtain the platform drift tolerance. In this embodiment, it is taken as 1.2kN.

[0079] If the absolute value of the difference between the stable load before and after the current verification does not exceed 1.2kN, the current dynamic suspension mass is used; if it exceeds 1.2kN, the current verification result is invalid and the verification is re-executed.

[0080] The dynamic quality working condition units of the air bucket are divided according to the support rope length, luffing position, benchmark lifting speed and verification acceleration level. The median value of the dynamic suspension mass formed by no less than 10 initial air bucket verifications within the same working condition unit is taken to obtain the dynamic quality benchmark of the air bucket.

[0081] Calculate the difference between the initial dynamic suspended mass of the air bucket and the dynamic mass benchmark of the air bucket, take the maximum absolute value of the difference, add twice the minimum resolution of the dynamic mass, and round up to an integer multiple of 10 kg to obtain the allowable boundary of the air bucket mass. In this embodiment, 180 kg is used.

[0082] The empty bucket dynamic mass benchmark is used to eliminate fixed dynamic deviations in paired verification; the empty bucket mass is the physical mass of the equipment, and the mass difference is obtained by subtracting the corresponding empty bucket dynamic mass benchmark from the current dynamic suspended mass.

[0083] When the weight difference is between -180kg and 180kg, the remaining suspended weight verification value is set to 0, the unloading confirmation status is set to empty bucket, and a normal return permit is generated.

[0084] When the weight difference is higher than 180kg but not higher than 250kg, the weight difference is determined as the remaining suspended weight verification value, the unloading confirmation status is determined as slight residue, and a normal return permit is generated.

[0085] When the mass difference exceeds 250kg, the mass difference is determined as the remaining suspension mass verification value. The slewing mechanism and luffing mechanism are kept at zero speed, and one re-unloading is performed.

[0086] When the quality difference is less than -180kg, the unloading confirmation status will be set to unconfirmed, no return permission will be generated, and the current result will not be used to update the dynamic quality benchmark of the empty bucket.

[0087] The re-unloading stroke is determined based on the net clearance height above the hopper, the safety distance of the grab bucket's outer contour, and the minimum lifting stroke required to cause relative displacement of the retained material. The re-unloading stroke must not exceed the available stroke after deducting the safety distance from the net clearance height. In this embodiment, the available stroke is higher than 0.30m, therefore the re-unloading stroke is taken as 0.30m. The re-unloading acceleration follows the low-impact verification acceleration of 0.20m / s². 2 .

[0088] After re-unloading, a complete verification is performed again. If the remaining suspended weight verification value after re-unloading is not higher than 250kg, a normal return permit is generated; if it is higher than 250kg but not higher than the upper limit of the low acceleration abnormal return weight, a low acceleration abnormal return permit is generated; if it is higher than the upper limit of the low acceleration abnormal return weight, no return permit is generated.

[0089] The upper limit of the return mass under low acceleration abnormal conditions is determined by the smaller of the mass calculated from the remaining load margin of the support rope under limited rotation and limited amplitude conditions, and the mass allowed to be carried away outside the hopper area. It is preferably 400 kg to 800 kg, and 600 kg is used in this embodiment.

[0090] After the low acceleration abnormal return permit takes effect, the upper limit of hoisting speed is 0.70 times the upper limit of normal return hoisting speed, the upper limit of hoisting acceleration is 0.50 times the upper limit of normal return hoisting acceleration, the upper limit of speed of slewing mechanism and luffing mechanism is 0.60 times the corresponding upper limit of normal return speed, and the upper limit of acceleration is 0.50 times the corresponding upper limit of normal return acceleration. The slewing anti-sway correction amount and the luffing anti-sway correction amount are included in the corresponding 0.50 times acceleration upper limit.

[0091] Each control upper limit is checked by substituting the initial return mass benchmark into the allowable torque of the actuator, braking deceleration, and available collision avoidance distance. If the upper limit of acceleration corresponding to the remaining torque of the actuator is lower than the proportional upper limit, or if the stopping distance calculated according to the current speed and braking deceleration exceeds the available collision avoidance distance, the lower upper limit formed by the check is adopted. In this embodiment, the normal return lifting speed is 0.35 m / s, and the upper limit of the lifting speed for low acceleration abnormal return is 0.25 m / s. Low acceleration abnormal return control continues until the delayed detachment is confirmed and a load mutation mass benchmark is formed, or the current return operation segment ends. In the preferred calculation example of task T0427, the dynamic suspended mass formed by the first verification is 8324 kg, the corresponding dynamic mass benchmark of the air bucket is 7805 kg, and the mass difference is 519 kg. Therefore, re-unloading is performed, and the dynamic suspended mass after re-unloading is 8143 kg. The remaining suspended mass verification value is 338 kg, which is higher than 250 kg and lower than 600 kg. Therefore, a low acceleration abnormal return permission is generated.

[0092] The unloaded weight of the grab bucket (7800 kg) is added to the remaining suspended weight (338 kg) to obtain the initial return weight baseline of 8138 kg.

[0093] S3 acquires return data, predicts the normal support rope driving torque based on the initial return mass benchmark, extracts the torque residual, reconstructs the suspended mass platform, and performs mechanical state competition by combining the changes in the suspended mass platform, the relationship between the drum rope length and the grab bucket displacement, and the periodic characteristics of the torque residual, generating one of the mechanical states among delayed detachment, wire rope slack, load oscillation, and unconfirmed state.

[0094] When the return permit takes effect and the hoisting mechanism begins its return journey, a return operation section is established. The position of the support rope suspension point is calculated based on the gantry crane's slewing angle, luffing position, and equipment geometric parameters, and the grab bucket position is converted to the same gantry crane coordinate system.

[0095] The direction of rope release from the drum is defined as the positive direction, and the direction of rope retraction is defined as the negative direction. The change in the length of the drum rope is calculated based on the difference between the current absolute rotation angle of the drum and the rotation angle of the drum at the start of the return stroke, the effective radius of the drum, and the side ratio of the grab bucket.

[0096] Calculate the length of each rope segment from the grab bucket to the suspension point of the support rope according to the rope winding structure and pulley position. After converting the length of each rope segment to the same grab bucket side ratio, add them together to obtain the tension rope length. Subtract the initial tension rope length of the return trip from the current tension rope length to obtain the change in tension rope length.

[0097] Dividing the equivalent load of the support rope by the equivalent axial stiffness of the support rope yields the elastic elongation of the support rope. Subtracting the initial elastic elongation of the return stroke from the current elastic elongation of the support rope yields the change in elastic elongation.

[0098] The rope length bearing residual is obtained by subtracting the changes in tension rope length and elastic elongation from the changes in drum rope length. The changes in drum rope length, tension rope length, and elastic elongation are all expressed in meters and use the same grab bucket side ratio.

[0099] The tension allowable belt is formed by no less than 10 confirmed tension return tasks. The maximum absolute value of the rope length bearing residual in each task is calculated, the maximum value at the task level is taken, and the equivalent length error of the drum position and the grab position is superimposed to obtain the half width of the tension allowable belt. In this embodiment, it is taken as 0.06m.

[0100] The dynamic equivalent mass of the wire rope is calibrated based on the initial empty bucket return mission, and the calibration units are divided according to the 2m support rope length range and three levels of negative acceleration, near-zero acceleration and positive acceleration.

[0101] The near-zero acceleration boundary is taken as twice the upper limit of the estimated noise level for lift acceleration. During the initial air-to-ground calibration, the absolute value of the estimated acceleration during the stationary and constant-speed operation phases should not exceed 0.025 m / s². 2 Therefore, -0.05m / s 2 Up to 0.05m / s 2 It is classified into near-zero acceleration levels.

[0102] For the initial empty bucket return sample, the equivalent load of the support rope is divided by the mass load converted acceleration, and then the empty mass of the grab bucket is subtracted to obtain the equivalent mass sample of the wire rope. The dimension of the mass load converted acceleration is m / s². 2 For no less than 5 equivalent mass samples of wire rope within the same calibration unit, the median value is taken to obtain the dynamic equivalent mass calibration node of the wire rope.

[0103] When the current support rope length is between two adjacent calibration nodes, linear interpolation is performed according to the distance ratio from the current length to the two nodes. When there are fewer than 5 samples in the current calibration unit and the calibration units with the same acceleration level within the adjacent 4m range, the suspended mass platform is not calculated, the mechanical state is set to unconfirmed, and only the wire rope tension protection is retained.

[0104] The grab position fitting window contains at least 6 grab spatial position feedback samples and does not exceed 1 / 8 of the theoretical swing period. In this embodiment, the grab spatial position feedback period is 50ms, and a 0.40s window is used, corresponding to 8 position samples.

[0105] The spatial position of the grab bucket is obtained by performing a quadratic curve fitting using the data from the current moment and 0.40 seconds prior. The first derivative of the fitted curve at the current moment is taken as the grab bucket velocity, and the second derivative is taken as the grab bucket acceleration. Position data after the current moment are not used.

[0106] Establish a unit direction pointing from the grab bucket to the suspension point of the support rope. Project the grab bucket's acceleration onto this direction to obtain the grab bucket's acceleration along the support rope. Add the product of gravitational acceleration and the cosine of the grab bucket's swing angle to the grab bucket's acceleration along the support rope to obtain the mass load-converted acceleration, in m / s². 2 .

[0107] The upper limit of the comprehensive error of the equivalent load of the support rope after compensation is formed by adding one times the load resolution to the maximum absolute deviation between the equivalent load of the support rope and the theoretical load in the initial empty bucket calibration. In this embodiment, the upper limit of the comprehensive error is taken as 0.8kN, and the allowable mass back-calculation error is taken as 180kg. The difference between the two is approximately 4.44m / s. 2 Therefore, the lower limit of the acceleration converted from mass load is taken as 4.5 m / s². 2 .

[0108] Before the delayed detachment is confirmed, the initial return mass benchmark of 8138kg is used as the current mass benchmark. After the delayed detachment is confirmed and the mass benchmark after the effective load change is formed, the mass benchmark after the load change is used as the new current mass benchmark.

[0109] The current mass benchmark is added to the dynamic equivalent mass of the wire rope, and then multiplied by the mass load conversion acceleration to obtain the normal support rope equivalent load. The normal support rope equivalent load is divided by the rope bearing capacity conversion factor to obtain the normal drum tangential force. The normal drum tangential force is multiplied by the effective drum radius to obtain the normal effective drum torque. The normal effective drum torque is added to the inertia compensation torque, Coulomb friction torque and viscous friction torque, and then divided by the transmission ratio and transmission efficiency to obtain the normal support rope driving torque.

[0110] The torque residual is obtained by subtracting the normal support rope driving torque from the measured support rope driving torque. Both the measured driving torque and the normal driving torque are expressed in N·m.

[0111] The support rope is kept taut, the rope length bearing residual is within ±0.06m, and the mass load converted acceleration is higher than 4.5m / s². 2 At that time, the instantaneous suspension mass is obtained by dividing the measured equivalent load of the support rope by the mass load to convert the acceleration, and then subtracting the dynamic equivalent mass of the wire rope.

[0112] The mass load converted acceleration is used to compensate for the gravity projection and the main body inertial load of the grab bucket along the support rope direction; the normal support rope driving torque is predicted using the current effective drum radius, static rope load conversion coefficient and transmission compensation parameters, and does not include the pulley wrap angle change caused by the lateral swing of the load, the multi-rope tension redistribution and the periodic change of transmission friction. Therefore, the periodic transmission additional load formed by the load swing is retained in the torque residual.

[0113] The suspended mass platform uses a 1.0s window median value and recursive smoothing, only used to determine whether the grab mass forms a continuous step change. Instantaneous suspended mass disturbances caused by short-term periodic oscillations are not written into the current mass benchmark. Load oscillation is confirmed only if there is no continuous mass step change within one theoretical oscillation cycle, the torque residual cycle recurrence, and the lateral velocity reversal. The instantaneous suspended mass within the most recent 1.0s is saved. The 1.0s window covers 20 grab space position feedback cycles and is less than 1 / 8 of the theoretical oscillation cycle corresponding to the 24.2m support rope. When there are no fewer than 10 valid instantaneous suspended mass samples and the sample time coverage is no less than 0.50s, the median value of the valid samples is taken to obtain the current window mass.

[0114] When the first effective window is formed, the current window quality is used as the initial value of the suspension quality platform. In subsequent cycles, the current window quality is subtracted from the suspension quality platform of the previous cycle, the difference is multiplied by the platform update coefficient, and then added to the suspension quality platform of the previous cycle to obtain the current suspension quality platform.

[0115] The platform update coefficient is calibrated between 0.30 and 0.70 to ensure that the platform fluctuation in the confirmed air combat mission does not exceed 180 kg, while ensuring that the known mass step change completes the platform transition within 1.5 s. In this implementation, 0.50 is used.

[0116] If there are fewer than 10 valid instantaneous suspended mass samples, or if the sample time coverage is less than 0.50s, maintain the suspended mass platform of the previous cycle, set the mechanical state to unconfirmed, and limit the upper limit of acceleration of the slewing mechanism and the luffing mechanism to 0.60 times the normal upper limit. Do not update the current mass benchmark during the unconfirmed state period.

[0117] The mechanical state competition first performs the wire rope loss judgment. The upper boundary of the loss protection range is formed by low-speed tension calibration. The positive torque of the hoisting drive is reduced step by step. The minimum positive torque at which the drive can still maintain continuous closed-loop adjustment is recorded and converted into the minimum controllable tension load according to the transmission relationship of S1. The larger value between the minimum controllable tension load and 3 times the equivalent load resolution of the support rope is taken as the upper boundary of the loss protection range.

[0118] When the equivalent load of the support rope is lower than the upper boundary of the tension loss protection range, and the rope length bearing residual exceeds the ±0.06m tension allowable zone for no less than two consecutive grab bucket spatial position feedback cycles, the wire rope is confirmed to be out of tension. In this embodiment, the grab bucket spatial position feedback cycle is 50ms, so the continuous tension loss confirmation time is taken as 100ms.

[0119] Once the wire rope is confirmed to be out of tension, stop the drum from continuing to release the rope. When the anti-collision, overload, and grab bucket position interlocks are all in the permissible state, control the drum to wind up the rope at a low speed. If the interlock conditions are not met, keep the drum stopped and output the tension protection waiting state.

[0120] The low-speed rope winding speed is the smaller of the minimum stable speed at which the hoisting drive can operate continuously in a closed loop and the upper limit of the re-tensioning impact speed. In this embodiment, the minimum stable speed of the drive is 0.05 m / s, so the low-speed rope winding speed is 0.05 m / s.

[0121] The tension recovery boundary is determined during the same low-speed tension calibration process that forms the upper boundary of the tension loss protection range: Starting from when the equivalent load of the support rope enters the tension loss protection range, the rope winding torque is gradually increased. The first load recorded when the equivalent load of the support rope continues to rise for five consecutive feedback cycles, the rope length catch-up residual re-enters the tension tolerance zone, and the hoisting drive speed fluctuation does not exceed 10% of the minimum stable speed, is taken as the minimum stable tension load. The larger value between the minimum stable tension load and the upper boundary of the tension loss protection range plus twice the resolution of the equivalent load of the support rope is taken as the tension recovery boundary, and it is required that the tension recovery boundary be lower than 20% of the static load of the empty bucket under the same working conditions. In this embodiment, the upper boundary of the tension loss protection range is 2.4 kN, the resolution of the equivalent load of the support rope is 0.2 kN, and the minimum stable tension load is 3.2 kN; therefore, the tension recovery boundary is taken as 3.2 kN. When the equivalent load of the support rope is higher than the tension recovery boundary of 3.2kN, and the rope length bearing the residual re-enters the tension allowable zone and continues for 6 grab bucket spatial position feedback cycles, the low-speed rope winding stops. In this embodiment, the tension recovery confirmation time is 0.30s. During the wire rope loss confirmation period, delayed detachment and load swing judgment are not performed.

[0122] If the wire rope is not confirmed to be out of tension, the negative deviation of the suspended mass platform relative to the current mass benchmark is calculated. If the negative deviation reaches 250kg and there are no new changes in the opening and closing of the buckets in the opening and closing mechanism commands and positions, a candidate state of mass reduction is formed, and the upper limit of acceleration of the slewing mechanism and the luffing mechanism is set to 0.60 times the upper limit of normal acceleration.

[0123] The 0.60x acceleration limit is determined based on the current operating speed, control delay, and available collision avoidance distance. The stopping distance is obtained by dividing the square of the current speed by twice the braking deceleration and adding the product of the current speed and control delay. The acceleration ratio is adjusted so that the stopping distance does not exceed the current available collision avoidance distance. This implementation adopts the 0.60x acceleration limit under the most unfavorable residual mass conditions.

[0124] Divide the current length of the tensioned rope by the acceleration due to gravity, take the square root of the quotient, and multiply it by 2π to obtain the theoretical oscillation period.

[0125] If, within one theoretical oscillation cycle after the formation of the candidate state of mass degradation, the suspended mass platform fails to recover to the ±180kg range of the current mass benchmark, the rope length bearing residual remains within the tension allowable range, and the opening and closing mechanism commands and positions remain unchanged, then a delayed detachment is confirmed.

[0126] After the delayed detachment is confirmed, the median value of the suspended mass platform within the following 1.0s is taken. If the difference between the maximum and minimum values ​​of the suspended mass platform within this 1.0s does not exceed 180kg, the median value is determined as the mass benchmark after the load change, and the mass benchmark after the load change is used as the new current mass benchmark.

[0127] After the load change and the mass reference stabilizes, the upper limit of acceleration of the slewing mechanism and the luffing mechanism is adjusted to 0.85 times the normal upper limit of acceleration. The 0.85 times acceleration upper limit is determined based on the fact that the sum of the basic acceleration and the maximum anti-sway correction acceleration does not exceed the allowable acceleration of the drive. When the grab bucket leaves the hopper protection area and does not form a new mechanical state for one consecutive theoretical swing cycle, the original upper limit of acceleration is restored.

[0128] If no mass degradation candidate state is formed, or if the mass degradation candidate state is formed and then disappears within one theoretical oscillation cycle and the suspended mass platform re-enters the ±180kg range of the current mass reference, the load oscillation confirmation is initiated; the suspended mass platform is only used to exclude continuous mass steps in this confirmation.

[0129] The periodic recurrence of torque residuals is calculated using effective data before and after the candidate state. If there is an effective torque residual of at least one theoretical oscillation cycle before the formation of the candidate state, the data from one theoretical oscillation cycle before the formation of the candidate state to one theoretical oscillation cycle after the formation is used. If the effective data before the formation is less than one theoretical oscillation cycle, the data from two theoretical oscillation cycles after the formation of the candidate state is used. Before the periodic window is formed, the mechanical state remains unconfirmed and is maintained at 0.60 times the upper limit of acceleration.

[0130] Subtract the window mean from the torque residual within the periodic window, and calculate the normalized correlation coefficient between the mean-residual torque and the sequence delayed by one theoretical oscillation period, and the normalized correlation coefficient between the mean-residual torque and the sequence delayed by half a theoretical oscillation period. Take the larger of the two absolute values ​​of the correlation coefficients as the periodic recurrence.

[0131] The periodic recurrence boundary is calibrated using confirmed load swing samples and confirmed non-swing samples. The midpoint between the minimum value of the periodic recurrence of the confirmed load swing samples and the maximum value of the periodic recurrence of the confirmed non-swing samples is taken. In this implementation, 0.72 is used. When the value ranges of the two types of samples overlap, automatic load swing confirmation is not enabled.

[0132] The load swing is confirmed when the suspended mass platform is within ±180kg of the current mass benchmark, the periodic recurrence is not less than 0.72, and the grab bucket lateral speed completes at least one directional reversal.

[0133] After confirming the load swing, the lateral displacement and lateral velocity of the grab relative to the suspension point of the support rope are decomposed into the slewing tangential direction and the amplitude radial direction, respectively. The lateral displacement feedback gain is multiplied by the displacement in the corresponding direction, and the lateral velocity feedback gain is multiplied by the velocity in the corresponding direction. The two products are added together and the opposite number is taken to obtain the initial corrected acceleration in that direction.

[0134] The slewing anti-sway correction acceleration is obtained by adding the slewing luffing coupling coefficient multiplied by the luffing initial correction acceleration; the luffing initial correction acceleration is obtained by adding the luffing initial correction acceleration multiplied by the slewing luffing coupling coefficient multiplied by the slewing initial correction acceleration. The two correction accelerations are each limited to within 30% of the upper limit of the normal acceleration of the corresponding mechanism, and then superimposed on the slewing and luffing basic acceleration commands.

[0135] The upper limit of the lateral displacement feedback gain is formed by dividing the maximum permissible anti-sway correction acceleration by the upper limit of the lateral displacement within the identification interval; the upper limit of the lateral velocity feedback gain is formed by dividing the maximum permissible anti-sway correction acceleration by the upper limit of the lateral velocity; the lower limits of the two gains are determined based on the condition that the anti-sway prediction model reduces the predicted amplitude to less than 50% of the initial amplitude within 3 theoretical oscillation cycles; and the upper limit of the gyration amplitude coupling coefficient is determined based on the condition that the corrected acceleration in both directions after cross-superposition does not exceed 30% of the upper limit of the normal acceleration.

[0136] After substituting the current equipment's lateral displacement, lateral velocity, and acceleration limits, the optimal lateral displacement feedback gain is 0.04s. -2 up to 0.16s -2 The preferred lateral velocity feedback gain is 0.15s. -1 up to 0.60s -1 The gyration amplitude coupling coefficient is preferably between 0.05 and 0.20, and the initial reference value in this embodiment is 0.08s. -2 0.35s -1 And 0.12.

[0137] When the grab bucket's lateral displacement is no higher than 0.15m, the absolute value of its lateral velocity is no higher than 0.05m / s, and it maintains this state for one theoretical oscillation cycle, it exits the load oscillation state. The exit boundary is formed by the grab bucket position detection error and the remaining oscillation amplitude allowed by the anti-sway control.

[0138] If the candidate state of mass degradation does not meet the late-onset detachment confirmation condition, and the load swing confirmation condition is not met after the suspended mass platform recovers, the mechanical state is set to unconfirmable state, and the 0.60 times acceleration upper limit is maintained without updating the current mass benchmark.

[0139] In the preferred calculation example of mission T0427, the initial mass benchmark of 8138kg is set as the current mass benchmark at the start of the return trip. After 1.7s of lifting during the return trip, the suspended mass platform drops to 7872kg, a drop of 266kg.

[0140] During this period, the rope length bearing residual was between 0.03m and 0.05m, within the ±0.06m tension allowable zone. The opening and closing mechanism did not produce any new actions. After approximately 9.9s of theoretical oscillation cycle, the suspended mass platform did not recover to the ±180kg range of 8138kg, confirming a delayed detachment. The median value of the subsequent stable platform, 7872kg, was determined as the mass benchmark after the load mutation.

[0141] S4 divides the prohibited inheritance interval and the continuous tension interval according to the mechanical state. Candidate anti-sway parameters are identified from the continuous tension interval. Under the same foundation suspension point acceleration, initial grab bucket state, support rope length sequence, current mass benchmark and external disturbance sequence, the closed-loop prediction model of candidate anti-sway parameters and benchmark anti-sway parameters is run respectively. The lateral displacement evaluation quantity and the support rope driving torque fluctuation evaluation quantity are compared, and the benchmark anti-sway parameter update and empty bucket load benchmark update are performed respectively.

[0142] The delayed detachment mass conversion interval is extended forward by the maximum data age of the data used for state determination, and continues until the mass reference stabilizes after the load change. In this embodiment, the delayed detachment determination involves grab bucket spatial position data, and its maximum data age is 100ms. Therefore, the delayed detachment mass conversion interval is extended forward by 100ms.

[0143] The maximum data age for the support rope driving torque and drum position data is extended forward in the wire rope untension range. In this embodiment, it is taken as 50ms and extended to 1.2s after the rope length bearing residual re-enters the tension tolerance zone. The 1.2s is formed according to the maximum value of the time required for the support rope load and rope length bearing residual to re-stabilize in the initial tension calibration.

[0144] The load swing entry and exit phases, the unconfirmed state interval, and the actuator interlock interval are all designated as prohibited inheritance intervals. The actuator interlock interval begins at the moment when any interlock in the collision prevention, overload, grab position, or drive state transitions from a permitted to a non-permitted state, extending forward by the maximum data age of the corresponding interlock data. After all related interlocks are restored to permitted status, this period continues until the corresponding feedback quantity re-enters the control allowable range and reaches the state recovery stabilization time; this moment is defined as the interval end. The state recovery stabilization time is determined by the effective maximum value of the time required from interlock restoration to the corresponding feedback quantity continuously and stably entering the control allowable range during initial equipment commissioning or post-maintenance calibration. If calibration records are insufficient, this data segment is not used for parameter inheritance.

[0145] By subtracting the prohibited inheritance interval from the return data, the remaining data simultaneously satisfy the following conditions: the rope is continuously tensioned, the rope length bearing residual is within the tension tolerance zone, the current quality benchmark remains unchanged, the protocol data is continuous, and the anti-sway control mode remains unchanged, thus forming a continuous tension interval.

[0146] Continuous tensioning intervals with a current quality benchmark difference of no more than 180kg, support rope lengths within the same 2m working condition unit, and the same anti-sway control mode are grouped into the same group.

[0147] The continuous tension interval should cover at least two theoretical oscillation cycles to ensure that the identification data includes complete positive and negative displacement and velocity changes. The anti-sway model includes three parameters to be updated. The continuous tension interval should include no less than four non-overlapping control change segments, of which three segments are used to provide independent parameter change directions and one segment is used as redundancy. If the conditions are not met, candidate anti-sway parameter identification will not be performed.

[0148] The anti-sway prediction model uses the slewing tangential displacement and velocity, and the amplitude radial displacement and velocity as state variables. The square root of the gravitational acceleration divided by the length of the support rope is used to obtain the natural angular frequency of the oscillation.

[0149] The damping coefficient of the fixed structure is calibrated through the free swing section in the initial return stroke of the bucket. A continuous interval is selected in which the slewing mechanism and the luffing mechanism maintain zero acceleration, the change in the length of the support rope does not exceed 0.10m, and the lateral swing of the grab bucket includes at least three peaks in the same direction. When the change in the length of the support rope does not exceed 0.10m, the change in the natural frequency of the swing caused by it is lower than the allowable change of the same 2m support rope length working condition unit. The three peaks in the same direction are used to form two independent attenuation intervals.

[0150] For two adjacent lateral displacement peaks in the same direction, calculate the natural logarithm of the ratio of the previous peak to the next peak, and then divide it by the time interval between the adjacent peaks to obtain the attenuation coefficient of the free swing segment. Take the median value of the attenuation coefficients of at least 5 effective free swing segments as the fixed structure damping coefficient for the corresponding support rope length working condition. When there are fewer than 5 effective free swing segments, use the previous effective version and do not update the candidate anti-sway parameters.

[0151] Within the parameter identification and parameter verification intervals, the basic slewing acceleration command, basic amplitude acceleration command, actual driver acceleration, and actual anti-slewing correction acceleration before anti-sway correction are superimposed are recorded simultaneously.

[0152] Based on the driver debugging data, an execution ratio is formed between the command acceleration and the suspension point acceleration. The base suspension point acceleration is obtained by subtracting the product of the actual executed anti-sway correction acceleration and the execution ratio from the actual suspension point acceleration. The base suspension point acceleration represents the mechanism motion input without anti-sway correction and remains the same during the closed-loop prediction process of candidate anti-sway parameters and reference anti-sway parameters.

[0153] In this embodiment, the same return state condition means that the closed-loop prediction process used to compare the candidate anti-sway parameters with the reference anti-sway parameters adopts the same initial grab state and keeps the support rope length sequence, the current mass reference and the external disturbance sequence consistent; the basic suspension point acceleration is formed by subtracting the actual anti-sway correction effect from the actual suspension point acceleration in the aforementioned manner, and the same sequence is used in the two types of closed-loop prediction processes.

[0154] For each control cycle, a linear fit is performed using the current and four previous measured lateral velocity samples. The slope of the fitted line is taken as the measured lateral relative acceleration. The external disturbance acceleration is obtained by adding the product of the measured lateral relative acceleration, the square of the oscillation natural angular frequency and the measured lateral displacement, the product of the fixed structure damping coefficient and the measured lateral velocity, and the actual suspension point acceleration. This external disturbance sequence is formed according to the control cycle order. The external disturbance sequence is used to characterize wind load, ship motion, and disturbances to unmodeled structures. It remains the same during the closed-loop prediction of candidate anti-sway parameters and baseline anti-sway parameters and is not recalculated based on the predicted state.

[0155] For any anti-sway parameter to be evaluated, the measured lateral displacement and measured lateral velocity at the beginning of the corresponding parameter identification interval or parameter verification interval are used as the predicted initial state. In each control cycle, according to the generation method of anti-sway correction acceleration in S3, a predicted anti-sway correction acceleration is generated based on the predicted lateral displacement and predicted lateral velocity corresponding to the anti-sway parameter to be evaluated, and the same acceleration limiting as the actual control is executed; the predicted suspension point acceleration is obtained by adding the result of multiplying the basic suspension point acceleration and the limited predicted anti-sway correction acceleration by the execution ratio.

[0156] For each direction in the slewing tangential and variable amplitude radial directions, the gravity recovery term, damping term, and predicted suspension point acceleration are subtracted from the external disturbance acceleration to obtain the predicted lateral relative acceleration. Using trapezoidal integral, the predicted lateral velocity is updated based on the average of the predicted lateral relative acceleration of the current cycle and the previous cycle, and then the predicted lateral displacement is updated based on the average of the predicted lateral velocity of the current cycle and the previous cycle.

[0157] The predicted swing angle is calculated based on the predicted lateral displacement and the length of the support rope. The centripetal acceleration is calculated based on the predicted lateral velocity. The product of gravitational acceleration and the cosine of the predicted swing angle, the lifting vertical acceleration, and the centripetal acceleration are added together to obtain the predicted mass load converted acceleration. The sum of the current mass reference and the dynamic equivalent mass of the wire rope is multiplied by the predicted mass load converted acceleration to obtain the predicted equivalent load of the support rope. The predicted support rope driving torque is then obtained according to the reverse transmission process in S3.

[0158] The parameters to be updated in the anti-sway model include lateral displacement feedback gain, lateral velocity feedback gain, and gyration amplitude coupling coefficient. The candidate parameter search range is determined to be ±8% of the baseline value based on the condition that the anti-sway correction acceleration formed by the candidate parameters does not exceed 30% of the upper limit of normal acceleration.

[0159] The search step size should ensure that the corrected acceleration difference formed by adjacent candidate parameters within the parameter identification interval is higher than the driver's acceleration command resolution. In this embodiment, a 4% parameter change satisfies this condition; therefore, the three parameters are taken at -8%, -4%, 0%, 4%, and 8% of the baseline value, respectively, forming a total of 125 sets of candidate parameters.

[0160] Each parameter identification and verification interval first uses the baseline anti-sway parameters for closed-loop playback. If the median absolute error between the predicted and measured lateral displacements corresponding to the baseline anti-sway parameters does not exceed twice the grab position detection resolution, and the median absolute error between the predicted and measured support rope driving torques does not exceed twice the support rope driving torque resolution, then the interval is determined as a valid closed-loop evaluation interval. If any error exceeds the corresponding boundary, the current interval is not used for candidate anti-sway parameter identification or verification.

[0161] Within the parameter identification interval for effective closed-loop evaluation, the closed-loop prediction model is run using each set of candidate anti-sway parameters and benchmark anti-sway parameters, under the same foundation suspension point acceleration, initial grab state, support rope length sequence, current mass benchmark, and external disturbance sequence. The median values ​​of the absolute values ​​of the predicted lateral displacement in the slewing tangential and variable amplitude radial directions are calculated, and the larger of the two median values ​​is taken as the lateral displacement evaluation quantity. The absolute deviation of the predicted support rope driving torque relative to the median value of the predicted support rope driving torque within the interval is calculated, and the median value of each absolute deviation is taken as the support rope driving torque fluctuation evaluation quantity.

[0162] Divide the lateral displacement evaluation value of the candidate anti-sway parameter by the lateral displacement evaluation value of the benchmark anti-sway parameter, and divide the support rope driving torque fluctuation evaluation value of the candidate anti-sway parameter by the support rope driving torque fluctuation evaluation value of the benchmark anti-sway parameter. Take the larger of the two ratios as the candidate evaluation value. When the evaluation value corresponding to the benchmark anti-sway parameter is lower than the corresponding detection resolution, use the corresponding detection resolution as the divisor, and determine the parameter combination with the smallest candidate evaluation value as the candidate anti-sway parameter.

[0163] From the continuous tension intervals in the same group, select one parameter identification interval and at least two parameter verification intervals that do not overlap with the parameter identification interval. A complete swing response enters only one of the intervals.

[0164] When there are no fewer than 5 stable return historical tasks under the same working condition, calculate the difference in lateral displacement evaluation and the difference in support rope driving torque fluctuation evaluation between adjacent historical tasks under the benchmark anti-sway parameters. Take the maximum value among the absolute values ​​of each type of difference and superimpose the corresponding detection resolution to form the upper limit of lateral displacement evaluation fluctuation and the upper limit of support rope driving torque evaluation fluctuation. When there are fewer than 5 stable return historical tasks, do not update the benchmark anti-sway parameters.

[0165] Within the entire parameter verification range, the lateral displacement evaluation quantity and the support rope driving torque fluctuation evaluation quantity corresponding to the candidate anti-sway parameter are both lower than the evaluation quantity corresponding to the benchmark anti-sway parameter. The reduction of the two types of evaluation quantities is higher than the corresponding evaluation fluctuation upper limit. The difference between the suspended mass platform and the corresponding current mass benchmark is within ±180kg. When the anti-sway correction acceleration generated by the candidate anti-sway parameter does not exceed the control limit, the candidate anti-sway parameter is qualified for updating.

[0166] When a candidate anti-sway parameter is eligible for updating, the update ratio is determined based on the upper limit of the allowable single-time change in the baseline parameter and the maximum allowable deviation of the candidate anti-sway parameter relative to the baseline anti-sway parameter. The product of the update ratio and the maximum allowable deviation must not exceed the upper limit of the allowable single-time change in the baseline parameter. The upper limit of the allowable single-time change in the baseline parameter is determined by the parameter change in adjacent valid parameter versions during the same stable return historical task, provided that control limiting is not triggered and both the lateral displacement evaluation and the support rope drive torque fluctuation evaluation do not exceed the corresponding evaluation fluctuation upper limit. In this implementation, the maximum allowable deviation of the candidate anti-sway parameter is 8%, and the calibrated upper limit of the allowable single-time change in the baseline parameter is 1.6%. Therefore, the update ratio is 0.20. The updated baseline anti-sway parameter is obtained by multiplying the baseline anti-sway parameter by 0.80 and the candidate anti-sway parameter by 0.20, and then adding them together.

[0167] If the candidate anti-sway parameter is not eligible for updating, continue to use the original benchmark anti-sway parameter, update the version record source task, quality benchmark working condition, supported rope length working condition, anti-sway mode and effective task, and start using it from the next unloading task, without modifying the control variables and status results already generated in the current task.

[0168] Task T0427 results in residual load and delayed shedding state, and does not enter the air bucket load baseline update set.

[0169] Subsequent independently confirmed empty bucket missions met the requirement that the remaining suspended mass verification value was 0, and that no delayed detachment, wire rope slack, or unconfirmable state occurred during the return trip. Furthermore, no residual material exceeding the corresponding resolution was detected in the bucket images and 3D point clouds. Each independently confirmed empty bucket mission generated one candidate value for the fully open bucket load.

[0170] In no less than 10 initial bucket calibration missions, the absolute difference between each fully open bucket load and its median value is calculated. The maximum absolute difference is taken and superimposed with 1 times the support rope equivalent load resolution to obtain the load consistency boundary.

[0171] After accumulating at least 5 fully open bucket load candidate values ​​for the same air bucket reference working condition unit, the median value of the candidate values ​​is calculated. When the absolute difference between each candidate value and the median value does not exceed the load consistency boundary, the candidate value enters the update set. If there are fewer than 5 candidate values, the original air bucket load benchmark continues to be used.

[0172] In a preferred calculation example, the five candidate values ​​for the fully open bucket load are 79.3kN, 79.5kN, 79.4kN, 79.4kN and 79.6kN, respectively, with a median value of 79.4kN.

[0173] The update ratio of the empty bucket load reference is determined based on the allowable single reference change amount, preferably 0.10 to 0.30. In this embodiment, it is 0.20. The original empty bucket load reference of 79.0kN is multiplied by 0.80, the median candidate value of 79.4kN is multiplied by 0.20, and the two are added together to obtain the updated empty bucket load reference of 79.08kN.

[0174] For confirmed air bucket tasks entering the update set, the difference between the current air bucket load and the original air bucket opening reference load in each of the 24 jaw plate opening intervals is calculated. When the signs of the differences in each opening interval are consistent and the absolute deviation between each difference and the difference in the fully open interval does not exceed the load consistency boundary, the change is determined as common mode load drift.

[0175] When the common mode load drift is established, the difference between the updated empty bucket load reference and the original reference is 0.08kN. The empty bucket opening reference load corresponding to each jaw plate opening degree in the same working condition unit is increased by 0.08kN to keep the shape difference of each opening degree unchanged.

[0176] When the common mode load drift is not valid, the original empty bucket load reference and the original empty bucket opening reference load are retained. The current working condition is marked as requiring re-execution of the empty bucket opening curve calibration, and no online write-back is performed.

[0177] The updated baseline anti-sway parameters and empty bucket load baseline record the parameter version, source task, applicable working condition and effective task respectively, and are used from the subsequent tasks. The update process does not modify the unloading release residual, remaining suspension mass verification value, return initial mass baseline and load change after mass baseline already formed in task T0427.

[0178] This invention also discloses an intelligent control system for port bulk cargo handling equipment based on protocol data, comprising: The protocol data acquisition module communicates with the hoisting drive, opening and closing drive, drum encoder, slewing drive, luffing drive and grab bucket position detection device to acquire grab bucket unloading protocol data and return data, and group them according to task identifier and unified control time scale. The unloading process analysis module forms the unloading operation section, performs transmission compensation on the drive torque of the support rope, establishes the load release trajectory according to the jaw plate opening, and generates the unloading release residual. The lifting verification module verifies the acceleration based on the unloading release residual configuration, controls the lifting mechanism to execute the reverse paired acceleration segment, and generates the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark. The return state recognition module generates rope length bearing residuals and torque residuals, reconstructs the suspended mass platform, executes mechanical state competition, and controls the hoisting mechanism, slewing mechanism, and luffing mechanism according to the mechanical state. The parameter inheritance module divides the prohibited inheritance interval and the continuous tension interval, identifies and verifies the candidate anti-sway parameters, and updates the benchmark anti-sway parameters and the empty bucket load benchmark respectively.

[0179] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent control of port bulk cargo handling equipment based on protocol data, characterized in that, include: Acquire grab bucket unloading protocol data, form unloading operation segments according to task identifiers and mechanical response intervals, perform transmission compensation on support rope drive torque, establish load release trajectory according to jaw plate opening, and generate unloading release residual. Based on the unloading release residual, the hoisting verification is initiated. When the slewing mechanism and the luffing mechanism are stationary and the hoisting direction remains unchanged, the hoisting mechanism executes the reverse paired acceleration segment. The paired torque response is processed by transmission conversion, length compensation and differential processing to generate the remaining suspended mass verification value, unloading confirmation status and return initial mass benchmark. Acquire return data, predict the normal support rope driving torque based on the initial return mass benchmark, extract the torque residual, reconstruct the suspended mass platform, and perform mechanical state competition by combining the changes in the suspended mass platform, the relationship between the drum rope length and the grab bucket displacement, and the periodic characteristics of the torque residual, generating mechanical states, including delayed detachment, wire rope loss of tension, load swaying, and unconfirmed state; in the case of delayed detachment, limit slewing and luffing and generate a mass benchmark after load change; in the case of wire rope loss of tension, stop rope release and perform tension protection; in the case of load swaying, generate slewing anti-sway correction and luffing anti-sway correction; Based on the mechanical condition, the system is divided into prohibited inheritance intervals and continuous tension intervals. Candidate anti-sway parameters are identified from the continuous tension intervals. The lateral displacement evaluation and support rope driving torque fluctuation evaluation are compared between the candidate anti-sway parameters and the benchmark anti-sway parameters under the same foundation suspension point acceleration and the same return state conditions. The benchmark anti-sway parameters are updated when both evaluation values ​​corresponding to the candidate anti-sway parameters are lower than those corresponding to the benchmark anti-sway parameters. The air bucket load benchmark is updated based on the fully open bucket load formed by multiple independently confirmed air bucket missions.

2. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 1, characterized in that, The specific operations for forming the unloading section are as follows: Based on the task identifier associated with the bucket opening control record and mechanism feedback record, the first moment when the position feedback of the opening and closing mechanism continuously changes along the bucket opening direction is determined as the mechanical response start point, and the moment when the jaw plate reaches the fully opened bucket position and the equivalent load of the support rope meets the load stability condition within the continuous sampling interval is determined as the mechanical response end point.

3. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 2, characterized in that, The specific steps for generating the unloading release residual are as follows: Based on the hoisting transmission chain parameters and drum motion state, the support rope driving torque is converted into an equivalent load of the support rope, and the difference is made with the empty bucket opening reference load that matches the support rope length, luffing position, and bucket opening speed. The resulting additional load is grouped into load nodes according to the jaw plate opening interval. Under the constraint that the load nodes do not increase with the overall jaw plate opening, the load release trajectory is reconstructed. The unloading release residual is generated based on the continuous stable load record after the bucket is fully opened.

4. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 1, characterized in that, The specific steps for performing the lifting verification are as follows: When the grab bucket is located in the safe verification zone of the bucket, the support rope is kept taut, and the slewing mechanism and the luffing mechanism are at zero speed, if the unloading release residual is not higher than the empty bucket residual boundary, the lifting verification is performed according to the verification acceleration corresponding to the empty bucket residual boundary; if the unloading release residual is not lower than the residual load boundary, the lifting verification is performed according to the verification acceleration corresponding to the residual load boundary; if the unloading release residual is located between the two boundaries, the verification acceleration is continuously configured according to the position of the unloading release residual between the two boundaries. When the lifting mechanism remains in the same lifting direction, it sequentially executes two acceleration segments with opposite acceleration directions and equal absolute values ​​of acceleration time integrals.

5. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 4, characterized in that, The specific steps for generating the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark are as follows: The torque responses of the lifting speed difference, support rope length difference and grab bucket swing angle difference in the two acceleration segments, which are respectively within the corresponding pairing allowable range, are reverse-paired. The average support rope length of each pairing sample is used as the pairing reference length. The paired torque responses are converted into load responses under the pairing reference length. The dynamic suspension mass is identified based on the obtained paired load responses and paired actual acceleration responses. When the absolute value of the difference between the stable load before and after verification does not exceed the platform drift allowance, the dynamic suspended mass is subtracted from the dynamic mass benchmark of the empty bucket under the corresponding working condition; when the absolute value of the obtained mass difference does not exceed the allowable boundary of the empty bucket mass, the remaining suspended mass verification value is determined to be zero; when the mass difference is higher than the allowable boundary of the empty bucket mass, the mass difference is determined to be the remaining suspended mass verification value; when the mass difference is lower than the negative number of the allowable boundary of the empty bucket mass, the unloading confirmation state is determined to be unconfirmable; when the absolute value of the difference between the stable load before and after verification exceeds the platform drift allowance, the unloading confirmation state is determined to be unconfirmable. When the unloading status is confirmed to allow the return trip, the sum of the unloaded mass of the grab bucket and the verification value of the remaining suspended mass is determined as the initial mass benchmark for the return trip.

6. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 1, characterized in that, The specific steps for extracting the torque residuals and reconstructing the suspended mass platform are as follows: Based on the actual rope winding structure of the support rope, the amount of rope released from the drum, the length of the tensioned rope path from the grab bucket to the suspension point of the support rope, and the elastic elongation of the support rope are converted to the same support rope ratio to form the rope length bearing residual. Based on the initial mass benchmark of the return trip, the dynamic equivalent mass of the wire rope that varies with the length of the support rope and the hoisting acceleration, and the acceleration of the grab bucket along the support rope direction, the equivalent load of the normal support rope is predicted. The equivalent load of the normal support rope is converted into the driving torque of the normal support rope according to the hoisting transmission chain relationship. The difference between the measured driving torque of the support rope and the driving torque of the normal support rope is calculated to generate the torque residual. When the support rope is kept taut and the rope length bearing residual is within the tension tolerance zone, the instantaneous suspended mass is formed based on the measured equivalent load of the support rope, the acceleration of the grab bucket along the support rope direction, and the dynamic equivalent mass of the wire rope. The suspended mass platform is then formed by recursively using the current and historical instantaneous suspended masses.

7. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 6, characterized in that, The specific operation for performing mechanical state competition is as follows: Before the delayed detachment is confirmed, the initial quality benchmark of the return trip is determined as the current quality benchmark; after the delayed detachment is confirmed, the quality benchmark after the most recent load change is determined as the current quality benchmark. When the equivalent load of the support rope enters the untension protection range and the rope length bearing residual continuously exceeds the tension allowable zone, the wire rope is confirmed to be untensioned, the drum is stopped from continuing to release the rope and tension protection is implemented. If the wire rope is not confirmed to be out of tension, and the suspended mass platform does not recover to the current mass benchmark within one swing cycle after descent, the rope length bearing residual is within the tension tolerance zone, and the opening and closing mechanism does not generate new bucket opening and closing actions, then the delayed detachment is confirmed. The stable suspended mass platform after descent is determined as the mass benchmark after the load change, and the acceleration of the slewing mechanism and the luffing mechanism is limited. If the wire rope is not confirmed to be out of tension, and the suspended mass platform does not form a step change that continuously deviates from the current mass benchmark within one swing cycle, the torque residual has swing cycle characteristics, and the lateral velocity of the grab bucket reverses direction, the load swing is confirmed, and the slewing anti-sway correction amount and the amplitude anti-sway correction amount are generated according to the lateral displacement and lateral velocity of the grab bucket. When delayed detachment, wire rope slack, and load oscillation cannot be mutually exclusively confirmed, the mechanical state is determined as unconfirmable, the restricted control of the slewing mechanism and luffing mechanism is maintained, and the current quality benchmark is stopped from being updated.

8. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 7, characterized in that, The specific steps for dividing the prohibited inheritance interval and the continuous tension interval are as follows: The delayed detachment quality conversion range, wire rope loss and retensioning range, load swing entry and exit range, unconfirmed state range and actuator interlock range are divided into prohibited inheritance ranges. The boundaries of the prohibited inheritance ranges are expanded according to the maximum data age of the data used to determine the corresponding mechanical state and the state recovery stabilization time. After deducting the prohibited inheritance interval from the return data, extract the continuous tension interval where the support rope is continuously tensioned, the rope length bearing residual is within the tension tolerance zone, the current quality benchmark remains unchanged, the protocol data is continuous, and the anti-sway control mode remains unchanged. Then, group the intervals according to the current quality benchmark, support rope length working condition, and anti-sway control mode.

9. The intelligent control method for port bulk cargo handling equipment based on protocol data according to claim 8, characterized in that, The specific steps for updating the baseline anti-sway parameters and the empty bucket load baseline are as follows: From the continuous tension intervals in the same group, select non-overlapping parameter identification intervals and parameter verification intervals, and identify candidate anti-sway parameters under the condition that the anti-sway model structure, equipment structure parameters and current quality benchmark remain unchanged. Candidate anti-sway parameters and benchmark anti-sway parameters are used respectively. Under the same basic suspension point acceleration, initial grab state, support rope length sequence, current mass benchmark, and external disturbance sequence, a closed-loop prediction model is run to generate lateral displacement evaluation quantity and support rope driving torque fluctuation evaluation quantity. When both evaluation quantities corresponding to the candidate anti-sway parameters are lower than the evaluation quantities corresponding to the benchmark anti-sway parameters, and the suspended mass platform is consistent with the current mass benchmark, the benchmark anti-sway parameters and candidate anti-sway parameters are weighted and synthesized according to the update ratio determined by the upper limit of the allowable single benchmark parameter change and the maximum allowable deviation of the candidate anti-sway parameters from the benchmark anti-sway parameters, and the benchmark anti-sway parameters are updated. Each independently confirmed air bucket mission generates a candidate value for fully open bucket load; after generating multiple candidate values ​​for fully open bucket load under the same working condition, the median value of the candidate values ​​is calculated. When the absolute difference between each candidate value and the median value does not exceed the load consistency boundary, the air bucket load benchmark for the corresponding working condition is updated.

10. A port bulk cargo handling equipment intelligent control system based on protocol data, employing the port bulk cargo handling equipment intelligent control method based on protocol data as described in any one of claims 1-9, characterized in that, include: The protocol data acquisition module communicates with the hoisting drive, opening and closing drive, drum encoder, slewing drive, luffing drive and grab bucket position detection device to acquire grab bucket unloading protocol data and return data, and group them according to task identifier and unified control time scale. The unloading process analysis module forms the unloading operation section, performs transmission compensation on the drive torque of the support rope, establishes the load release trajectory according to the jaw plate opening, and generates the unloading release residual. The lifting verification module verifies the acceleration based on the unloading release residual configuration, controls the lifting mechanism to execute the reverse paired acceleration segment, and generates the remaining suspended mass verification value, unloading confirmation status, and return initial mass benchmark. The return state recognition module generates rope length bearing residuals and torque residuals, reconstructs the suspended mass platform, executes mechanical state competition, and controls the hoisting mechanism, slewing mechanism, and luffing mechanism according to the mechanical state. The parameter inheritance module divides the prohibited inheritance interval and the continuous tension interval, identifies and verifies the candidate anti-sway parameters, and updates the benchmark anti-sway parameters and the empty bucket load benchmark respectively.

Images