Multi-channel rope paying-in and paying-out differential loading and length measuring device and method

Abstract

The present application relates to a kind of multi-channel rope winding and unwinding difference loading and length measuring device and method.Each channel detects the motion of the rope storage wheel on the output side of the magnetic powder clutch, determines the length of the rope storage and the effective loading arm respectively with the same rope storage state and calibration version, calculates the length of the winding and unwinding rope and adjusts the load in combination with the target tension and the measured tension.The position and tension sample of the necessary channel are checked for timeliness, time deviation, state and version correspondence, and the single tension error and channel load difference error are verified;After continuous compliance and reaching the observation time, the load condition mark is associated with the corresponding length, state and version record, and the non-compliant and transitional data are retained.The present application is used to identify the length record in the difference loading test that is not suitable as the compliant working condition result due to state mismatch or load condition deviation.

Description

Technical Field

[0001] This invention relates to loading tests and length measurements of rope reeling and releasing mechanisms, specifically to a device and method for linking rope length measurement, magnetic powder loading control, and load condition verification under multiple different tension conditions by utilizing the rope storage state. Background Technology

[0002] The rope winding and unwinding test equipment can form a loading loop using a drive motor, magnetic powder loading device, reel, and tension sensor. As the amount of rope on the reel changes, the effective lever arm of the rope acting on the reel also changes. Existing technologies reduce the impact of this change on the loading through tension feedback, and there are also schemes that use a combination of feedforward and feedback control based on the reel diameter information.

[0003] For example, CN104502133A discloses a winch constant load control test platform; CN120903330A discloses a winding control system combining multi-axis magnetic powder devices, winding diameter information, and tension feedback. For multi-channel differential loading tests, it is also necessary to establish the correspondence between length records and actual load conditions.

[0004] In multi-channel differential loading tests, the slippage of the magnetic powder clutch prevents the input-side motion from directly representing the rope storage wheel motion; changes in the amount of rope stored further alter the length calculation relationship and the loading lever arm with the state. If the position and tension are obtained at different times, or if different calibration versions are used for length and lever arm calculations, even if each displayed value is within a reasonable range, inconsistent length and load records may still occur. Furthermore, meeting the single-channel tension error requirements does not necessarily guarantee that the load difference between channels meets the requirements. Therefore, it is necessary to ensure that the output-side rope storage state, loading calculation, and load condition verification results correspond to each other and to check their consistency. Summary of the Invention

[0005] The technical problem solved by this invention is: in rope winding and unwinding tests with magnetic powder clutch slippage, rope storage variation, and multi-channel sampling time difference, the length measurement results marked as meeting the working conditions correspond to the effective rope storage state under the same calibration system and the sampling window that meets the predetermined single-path tension and inter-channel load difference conditions.

[0006] This invention discloses a multi-channel rope take-up and release differential loading and length counting device, comprising a controller, a recording unit, and at least two rope take-up and release channels. The drive and deceleration units of each channel are connected to the rope storage wheel via a magnetic powder clutch; an encoder is located on the output side of the magnetic powder clutch and is coaxially connected to the rope storage wheel or connected through a mechanism with a known transmission ratio, used to obtain the position of the rope storage wheel; a tension sensor detects the rope tension in that channel. The controller collects the position and tension of each channel and outputs loading commands to the corresponding magnetic powder controller.

[0007] The controller uses the initial rope storage volume determined after installation as a baseline, updates the rope storage status of each channel based on signed position increments, and fixes the status, channel identifier, and calibration version of the same channel for the same update, forming a status snapshot. Length counting and load calculation respectively query the rope storage length relationship and effective loading lever arm relationship from this snapshot, calculating the rope extension / retraction length based on the difference between the initial rope storage length and the current rope storage length. The two calibration relationships establish nodes based on the number of rope turns and interpolate within the valid range; calculation results outside the range are not published as valid results.

[0008] The effective loading lever arm refers to the equivalent vertical distance from the line of action of the rope tension to the selected torque calculation axis. The length calculation relationship and the effective loading lever arm relationship are determined according to the length measurement and torque conversion requirements, respectively. The relationship between the output control quantity and the actual transmitted torque is calibrated separately to distinguish between geometric quantities and actuator characteristics.

[0009] At least two channels operating simultaneously employ different target tensions. Each channel generates a feedforward torque based on the target tension and the effective loading arm, superimposed with feedback correction for the measured tension error, and converts this into a magnetic powder control quantity via actuator calibration, with amplitude and rate of change limits applied. Each channel independently stores its integral state and the previous control quantity, and performs loading calculations independently; the length measurement and loading reference for this channel both use the same rope storage state.

[0010] The controller checks the timeliness of the sampling data, acquisition time deviation, and the correspondence between status and version for the position and tension samples of each necessary channel specified in the current test formula. Candidate calculation results are submitted uniformly after all necessary channels have passed the check; if a check fails, the original sample and the reason for the failure are retained. Single-channel tension error is taken as the difference between the measured tension and the target tension; channel-to-load difference error is taken as the difference between the measured tension difference and the target tension difference, and both are verified according to the set tolerances.

[0011] When the error, data correspondence, and control constraints continuously meet the set conditions and the observation period is reached, a load condition compliance mark is generated and saved in association with the corresponding length measurement result, rope storage status, and calibration version. Simultaneously, records of non-compliance and transitions are retained. The original observation window is terminated when the formula or calibration version changes, the direction is reversed, or an invalid sample appears; subsequent samples meeting the conditions are re-established in a new window, not spliced ​​with the terminated window.

[0012] This method is executed in the order of data collection, status update, calculation, inspection, verification, and recording described above. For specific calculation relationships, please refer to the implementation method.

[0013] Compared with the prior art, this application has the following advantages: The output-side position detection reflects the movement of the rope storage wheel, avoiding the use of the input-side rotation amount containing the slip of the magnetic powder clutch as the rotation amount of the rope storage wheel; the initial rope storage amount and the signed position increment provide the same rope storage reference for the round trip length measurement.

[0014] Using the same state snapshot and calibration version for both length measurement and loading reference reduces calculation errors caused by asynchronous state updates or mixed calibration versions. Length relationships and effective loading lever arm relationships are calibrated separately, ensuring that each calculation uses its corresponding measurement basis.

[0015] Each channel calculates the loading command based on the independent target tension, effective loading lever arm, and measured tension, forming multi-channel differential loading conditions; the length record is associated with the load conditions, making it easy to compare the rope release and take-up performance of each channel according to the working conditions.

[0016] Compared to verifying only the tension of a single channel, adding channel-based load difference error verification can identify samples where the individual channel errors are all within tolerance but the load difference between channels has exceeded the tolerance. Compared to verifying only the load difference, single-channel error verification can identify samples where all channels have a common offset but the load difference remains unchanged. The compliance determination for both types of verification under the common constraint differential loading condition is then performed.

[0017] By checking the timeliness of the sampling data, the deviation in collection time, and the correspondence between status and version, expired samples, samples with excessive collection time deviation, or status and version mismatches can be excluded. Observing for a set duration and re-establishing the window after switching operating conditions, reversing directions, or abnormalities can distinguish between continuous compliance segments and transitional processes. Saving window start and end information, non-compliance records, and transition records facilitates verification of the load conditions corresponding to the length calculation results of each segment.

[0018] The load condition markings only describe the sampled window under test; the accuracy of physical length measurement and the dynamic error of tension still depend on the measurement, calibration and control performance. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the functional connection of a four-channel device.

[0020] Figure 2 This is a flowchart for status snapshot, length measurement, loading, and load condition verification.

[0021] Figure 3 A schematic diagram illustrating the creation and reconstruction of the load condition observation window.

[0022] Reference numerals: 10-Controller; 11-Drive and deceleration unit; 12-Magnetic powder clutch; 13-Rope storage pulley; 14-Encoder; 15-Tension detection unit; 16-Test interface. Detailed Implementation

[0023] like Figure 1 As shown, each channel is equipped with a drive and deceleration unit 11, a magnetic powder clutch 12, a rope storage wheel 13, an encoder 14, and a tension detection unit 15. The rope is connected to the test interface 16 via a guide. The controller 10 acquires the position, tension, and valid flag of each channel and independently saves the control status of each channel. Figure 1 All signal and power lines are omitted. The encoder 14 detects the movement of the rope storage pulley 13 on the output side of the magnetic powder clutch 12, directly replacing the rotation of the reel without rotating the motor input side.

[0024] After installing the rope, record the encoder 14 count and the actual rope quantity at the known rope storage position. Reconfirm this correspondence after reinstalling the rope, changing the encoder 14 zero point, or data loss. The initial rope quantity is the amount of rope stored on the rope storage reel 13 at initialization, and it needs to be correlated with the encoder 14 count at that moment. The four-channel configuration is only an example; the number of channels is determined by the actual test interface, and each channel can be configured with its own allowable load.

[0025] The magnetic powder clutch control quantity adjusts the transmitted torque amplitude; the actual torque direction also depends on the relative motion of the input and output and the transmission relationship. When the rope storage pulley 13 is driven by an external mechanism to release and actively retrieve the rope, corresponding driving conditions are adopted respectively. Analog quantities are used to adjust the transmitted torque amplitude; the torque direction is determined by the relative motion of the input and output and the transmission relationship. Mechanical braking, emergency stop, and drive permission are verified independently of load conditions.

[0026] The encoder 14 is coaxially connected to the rope storage wheel 13 or connected via a known transmission ratio, and the count per revolution is uniformly converted to the rope storage wheel shaft. The magnetic powder clutch allows for a speed difference between the input and output, therefore the cumulative rotation angle on the input side cannot be converted into the number of rope turns based on a zero-slip transmission relationship. When output-side detection is used, the length count and load reference are updated by the actual rotation of the rope storage wheel 13, rather than being inferred from the drive command.

[0027] One embodiment of single-row radial overlay uses a rope with a nominal diameter of 1.6 mm and a rope storage wheel 13 with a nominal groove width of 1.7 mm, with the groove walls constraining the axial parallel winding. This dimensional combination is only for illustrating a structural embodiment; manufacturing tolerances, rope compaction, and arrangement should meet the applicable conditions of the calibration relationship used. The calibration nodes must cover the actual rope storage range used.

[0028] In a four-channel embodiment, the drive and loading components for each channel are centrally mounted on the same equipment base. The rope storage wheel 13, rope guide, encoder 14, and tension detection unit 15 are arranged in channels. The control cabinet is equipped with a magnetic powder controller and measurement interface corresponding to each channel. The control terminal displays the target tension, measured tension, rope storage length, and position or speed status for each channel, and provides single-channel operation and multi-channel simultaneous operation functions. This arrangement allows different channels to use different target tensions during the same test period, and enables the operator to correlate the tension display with the length counting status during the same period for observation.

[0029] The above display and structural arrangement are intended to illustrate a feasible device configuration and do not limit the number of channels, controller model, or display interface format. Displayed values ​​are state quantities at the time of sampling; they are only used for load condition verification when the sampling time, recipe version, calibration version, and state index have all been checked and recorded. Continuous operation accuracy needs to be determined separately.

[0030] Status Update: (1) In the formula, Channel number, The sampling sequence number; The original position count is a dimensionless integer; This represents the original difference between adjacent counts.

[0031] Let the counting modulus of encoder 14 be M. When the original difference is greater than half of M, subtract M; when it is less than half of negative M, add M; otherwise, retain the original difference to obtain the unwinding increment. The sample is rejected if it is exactly at the positive or negative half-range. This process requires that the maximum possible change in count between adjacent valid samples is less than half of the counting modulus, and ensures unambiguity through sampling time limit and maximum displacement checks. (2) In the formula, This is the cumulative signed count since initialization, with an initial value of 0; To set the installation direction symbol, use 1 or -1 to increase the amount of rope stored corresponding to the received rope. The increment of the count after unwinding; (3) In the formula, To store the number of rope loops, The subscript 0 indicates the initial value; This is to calculate the count per revolution of the storage rope reel 13; The direction symbol is already included, so the direction will not be multiplied again; the number of rotations is not zeroed at the 360-degree boundary.

[0032] The difference between adjacent counts is based on the original count saved from the previous valid sample, and its acquisition time is retained; invalid samples do not update this baseline. If the change in count cannot be guaranteed to be less than half of the counting modulus during data loss, or if the initial rope quantity cannot be confirmed after power-on, the system remains in a pending confirmation state; updates will only resume after the correspondence between count and rope quantity is re-established at a known rope location.

[0033] Relationship between length and effective lever arm: Establish node groups indexed by the number of rope coils for each channel. Each node includes the number of coils, rope length, and effective loading lever arm, along with the channel identifier, calibration version, and effective range. The number of coil nodes and length nodes are strictly incremental, and the lever arm is a positive value. Nodes can be set at integer or non-integer coil positions; additional nodes can be added for rapidly changing intervals. (4) In this formula, the channel and sampling index are omitted. Current lap number The index of the node with the smaller number of cycles between the nodes at both ends of the interval; and These represent the number of cycles at the two ends of the interval, respectively. This is a dimensionless interpolation ratio, located between 0 and 1; (5) In the formula, This represents the current length of the stored rope. and The length of the rope stored at the corresponding node, in meters (m). Calculated according to equation (4), linear interpolation is used between nodes; (6) In the formula, The current effective loading arm, and The lever arm of the corresponding node, in meters (m); Calculated according to formula (4), this relationship uses the same state and calibration version as the length relationship; (7) In the formula, The net amount of rope released relative to the initial state, in meters, with release being positive and income being negative; For the first Relationship of rope lengths in the channel; and These are the initial and current lap numbers, respectively. (8) In the formula, The increment of rope release between adjacent valid samples, in meters; The previous valid sample number is 1000 laps, and the other parameters are the same as in equation (7).

[0034] The lengths mentioned above are geometric lengths measured as the rope enters and exits the reel, and do not directly include elongation under stress, slack, or external slippage. When the number of turns returns to its original value, the calculated length under the fixed calibration also returns to its original value. If rope compaction, slack, or rearrangement invalidates the calibration, the length measurement result will be marked as invalid. Length calculations are limited to the valid calibration range.

[0035] Calibration implementation method: Under controlled installation and pre-tensioning conditions, the amount of rope entering and exiting the reel at each node is measured using an independent length reference, and the length node is obtained from the reference rope storage amount. The effective lever arm can be determined by the vertical distance from the reel axis to the tension line of action and the known transmission relationship, and checked under selected working conditions; if torque and tension are used for joint identification, friction, inertia, and transmission losses should be deducted or modeled separately. The correspondence between the simulated output and the transmitted torque is established through calibration. The applicability of the calibration relationship is checked separately under the rope take-up and release working conditions; if both working conditions meet the calibration requirements, the common relationship is used; otherwise, the corresponding calibration relationship is re-established.

[0036] The length and effective lever arm nodes form a calibration package for the same channel. Before publishing the calibration package, check the node order, range, and units. After publishing, it remains unchanged within a test window. Modifying any relationship will create a new version, and nodes will not be covered item by item in ongoing calculations. For compaction or friction effects related to the extension / retraction direction, first determine whether it falls within the applicable range of the shared calibration package. If it does not, terminate the current window and reconfirm.

[0037] The independent length references, force measurements, and shaft-side transmission relationships required for calibration should have sufficient records to reproduce the node establishment process. When the tension sensor is installed in the force path of the guide wheel, the sensor output should be converted into rope tension according to the actual wrap angle and force direction, or the complete installation assembly should be calibrated directly. The verification tolerance is determined based on the test requirements and measurement uncertainty.

[0038] Difference loading calculation: (9) In the formula, The feedforward torque is converted to the storage rope pulley axle, in N·m; The target tension of this channel, in N; The effective loading arm of this channel is expressed in meters (m). , For channel and sampling sequence number; (10) In the formula, Tension error used for load control, in nanometers (N); The target tension is F; the measured tension corresponding to the current state is in N. (11) In the formula, These are the candidate values ​​for this score. The torque of the previous integral is expressed in N·m. The integral coefficient is expressed in m / s. The sampling interval is in seconds (s). Calculate according to formula (10); (12) In the formula, This is the unlimited requested torque, in N·m; This is a proportionality constant, in meters (m). This is the integral torque used in this test, in N·m; and Calculate according to formulas (9) and (10) respectively.

[0039] Based on the actuator's calibration under the corresponding operating conditions, the requested torque is converted into a magnetic particle control quantity. This control quantity is limited by the channel rating, the thermal capacity of the magnetic particle device, the controller's range, and its rate of change. If using the current integral candidate value would cause the control quantity to further exceed the amplitude or rate of change limits, the previous integral value is retained, and the output is recalculated. Both amplitude and rate of change limits must be included in the integration process.

[0040] If a linear torque-voltage relationship can be used to approximate the voltage within a small, calibrated range, the following formula can be used for conversion; when the effective range of this approximation is exceeded, the actual calibration curve should be used: (13) In the formula, Unlimited voltage request, unit: V; The torque voltage gain referred to the sheave shaft, in N·m / V, must be positive and calibrated; Calculate according to formula (12).

[0041] Instructions for all magnetic particle controllers are written uniformly from the same output processing module. Different channels use independent input, output, and integral states. When modifying the load formula, first check whether the settings of each channel are within the allowable range. After all checks pass, update the formula version. The updated formula remains unchanged until the next valid modification.

[0042] Status snapshot fixation and associated result submission: like Figure 2At the start of each correlation calculation, the controller locks the current load formulation and calibration packages for each channel, collecting output-side position, tension, acquisition time, and data validity indicators. Each channel first verifies the input, then prepares candidate values ​​for rope status, length, effective lever arm, and loading reference; no new results are released for any channel before the check is complete. After each required channel passes the data correspondence check, a common correlation number is assigned, and candidate results are released. The number of rotations, lever arm, and integral status of each channel are independent; shared status means that length counting and loading use the same status within the same channel. Required channels are the set of channels that must be simultaneously verified according to the current test formulation, fixed within an observation window; changes in the set are treated as formulation changes. Control constraints include channel amplitude range, rate of change, and actuator operating permission; samples with amplitude limits, rate of change limitations, or actuator operating permission not met are not included in the load condition compliance window.

[0043] Acquisition time refers to the actual moment when position or tension data is obtained, using a unified or calibrated time base. The timeliness of the sampled data is judged by the time interval between the current inspection time and the acquisition time: this interval should be between zero and the allowable time limit; samples exceeding this range are considered invalid. When position and tension data are time-stamped separately, the acquisition time difference between the two on the same channel, as well as the acquisition time deviation between each necessary channel, is checked. For equipment without acquisition timestamps, a verified synchronous acquisition mechanism can be used, incorporating conversion and transmission delays into the time deviation and timeliness checks. In case of inspection failure, the original sample and the reason are saved, and the sample is marked as invalid.

[0044] The measured tension collected in this study is used for error verification and feedback calculation for the next loading command. The tension change caused by this loading command is obtained from subsequent sampling. The record distinguishes between the sampling sequence number, the command issuance sequence number, and the constraint flag; the load condition judgment is based on the actual tension obtained. If the data is invalid, the update of the loading command based on that sample is suspended, and the predetermined hold, unload, or stop actions are executed by independent safety control logic.

[0045] Load condition verification and recording: (14) In the formula, To verify the single-path tension deviation used, in nanometers (N), and the control error. Opposite signs; and These are the actual and target tensions, respectively. , For channel and sampling sequence number; (15) In the formula, , For different channels, The measured load difference is the deviation from the target load difference, in N; F is the actual tension. Target tension, in N; For the corresponding sampling sequence number; (16) In the formula, For channel Permissible tension error The allowable load difference error for the selected channel pair, in N; and The thresholds are calculated according to equations (14) and (15) respectively, and the thresholds are determined by the demand and measurement uncertainty.

[0046] For samples under the same load formulation and calibration version, check the timeliness of the sampling data, the time difference between the acquisition of the same channel position and tension, the acquisition time deviation between channels, and the sampling sequence number. If the data is expired, the time deviation exceeds the limit, the control quantity reaches the limit condition, or the status and version are not compatible, save the corresponding reasons, and the sample will not enter the load condition compliance window.

[0047] When samples that meet the conditions of equation (16) and data validity are received continuously and the set observation time is reached, the sampling window can be marked as meeting the load conditions. When the formula changes, the calibration version changes, a reversal is detected, or an invalid sample appears, the observation window is re-established; when a stationary sample is encountered, the previous non-zero motion direction is retained. This mark only describes the sampling window under test and does not directly determine whether the tested mechanism is qualified, nor does it prove that the conditions are met at every moment within the sampling interval.

[0048] like Figure 3 Window status includes no matching window, under observation, and matching window. The first matching sample only establishes a starting point; subsequent samples must also meet the maximum adjacent interval. A matching window marker is generated only after the observation period is reached. When the formulation, calibration, orientation, or validity conditions change, the original window start and end numbers and corresponding length range are saved, and the window is closed; subsequent timing starts from the new first matching sample. Multiple non-contiguous windows can be included in the report separately, but cannot be combined into a single continuous matching condition.

[0049] Records of compliance, non-compliance, transitions, and anomalies during the test process are saved, including original location, target tension, measured tension, time, status, calibration version, output limit marker, and reason for non-compliance. The hold, load reduction, or stop actions following any anomaly are executed by independent safety control logic according to a predetermined strategy.

[0050] Threshold relationship: Since the channel pair deviation equals the difference between the two channel deviations, its absolute value does not exceed the sum of the absolute values ​​of the two channel deviations. If the channel pair tolerance is not less than the sum of the two single-channel tolerances, and the tension errors of both channels are within their respective tolerances, the channel pair inspection will not impose a more stringent constraint. Therefore, the channel pair tolerance should be selected in conjunction with the test requirements; when an additional difference constraint is required, it should be less than the sum of the two single-channel tolerances.

[0051] Calculation examples and implementation constraints: The following calculation process is illustrated using set parameters; the values ​​are not actual machine calibration or test results. For a certain channel node: loop number 1 corresponds to a rope length of 0.32m and a lever arm of 0.0516m; loop number 2 corresponds to a length of 0.65m and a lever arm of 0.0532m. At the current loop number of 1.5, interpolation yields a length of 0.485m and a lever arm of 0.0524m. If the initial loop number is 2, the net rope release is 0.165m. With a target tension of 100N, the feedforward torque is 5.24N·m.

[0052] The target tensions for the four channels are set to 100N, 50N, 30N, and 10N, used only to illustrate differential formulations. If the actual tensions are 103N, 53N, 33N, and 13N respectively, the target differences between each channel are maintained, but each channel is 3N higher than the target; checking only the load difference will miss the common deviation. If the actual values ​​of the first two channels of another sample are 101N and 49N, the single-channel deviations are 1N and -1N respectively, and the load difference deviation between the two channels is 2N; with a single-channel tolerance of 2N and a difference tolerance of 1N, the single-channel check passes, but the difference check fails.

[0053] Equipment operating parameters need to be determined separately according to the rated capacity and mechanical limitations of each channel. Comparison and verification can be performed while maintaining the same mechanism, rope, initial rope quantity, and motion conditions, comparing the existing control with this implementation method for tension error, load difference error, length reference error, and data correspondence anomalies.

[0054] This method does not require the permanent addition of a torque sensor. During the calibration phase, independent measuring tools can be used, and the tool accuracy, shaft-side conversion relationships, measurement conditions, and the correspondence between control quantities and transmitted torque should be recorded. Equipment operation should utilize confirmed actual calibration relationships, corresponding versions, and valid rope storage status.

[0055] On-site display and verification calculation of differential formulations: In a single-frame interface recording of a four-channel loading device, the target tensions for channels A, B, C, and D are displayed as 330N, 50N, 50N, and 50N, respectively, with actual tension values ​​of 329.944N, 49.868N, 49.998N, and 41.589N. Based on these displayed values, the single-channel deviations are -0.056N, -0.132N, -0.002N, and -8.411N, respectively. The actual display difference between A and D is 288.355N, while the target difference is 280N, a difference of 8.355N. In another single-frame recording, the target tensions are 330N, 330N, 50N, and 50N, with displayed values ​​of 330.186N, 330.781N, 49.802N, and 40.909N. The display deviation for channel D is -9.091N.

[0056] The above records illustrate the actual scenarios targeted by different load formulations and deviation checks. These are interface display records and cannot replace independent calibration, nor can they be used to determine the four-way closed-loop status or continuous loading accuracy. Without obtaining the corresponding length of acquisition time and status version, the tension and length of different images will not be stitched together to form a single load condition compliance sample.

[0057] Offline numerical and logical verification examples: An offline calculation program, independent of the device's inputs and outputs, was used to perform 41 numerical and logical tests on rope status updates, the relationship between rope length and effective loading lever arm, loading calculations, and load verification. All 41 tests passed. The inputs used manually set values ​​and calibration nodes, without connecting to a PLC, actuator, or actual sensors. The verification scope was limited to program branches and calculation rules, and did not involve actual machine performance testing.

[0058] In the verification of equations (14) to (16), the four target values ​​are set to 100N, 50N, 30N, and 10N, the single-channel tolerance is set to 2N, and the channel pair tolerance is set to 1N. When the inputs are 103N, 53N, 33N, and 13N, the difference deviations of the six channel pairs are all 0, but the single-channel deviations of the four channels are all 3N, and the double verification rejects the sample. When the inputs are 101N, 49N, 30N, and 10N, the single-channel deviations of the four channels do not exceed 2N, while the load difference error between the first and second channels is 2N, and the double verification also rejects the sample. Compared with the rules that only check single channels or only check differences, these two deterministic use cases illustrate the deviations that may be missed when using the two verification rules alone.

[0059] Taking another sample where the tension exactly matches the target, the allowable channel time deviation is set to 20ms, the receiving time to 0.10s, and the acquisition times for each channel to 0.10s, 0.06s, 0.10s, and 0.10s. Although the tension error is 0 for all samples, the sample does not enter the matching window because the maximum acquisition time difference is 40ms. Changing the calibration version used for loading calculations in a channel to a version different from the calibration version saved in the channel's state snapshot also results in rejection. This test case is used to check the time and version verification logic; the input time and version are both offline settings.

[0060] The maximum sample interval and observation duration for the observation window test case are 0.15s. When matching samples are received consecutively at 0.10s, 0.20s, and 0.30s, the cumulative observation time from the first to the third sample is 0.20s, reaching the set observation duration, and the window meets the conditions. The window terminates when a tension sample exceeding the tolerance appears at 0.40s, and only a new starting point is established after the matching samples are restored at 0.50s. All five samples are retained; matching samples before and after an anomaly cannot be spliced ​​into a continuous matching result. The above thresholds are for logical reproduction only; the actual values ​​should be determined based on the equipment sampling performance, test requirements, and measurement uncertainty.

Claims

1. A multi-channel rope take-up and release differential loading and length counting device, comprising a controller and at least two channels, each channel being provided with a drive unit, a magnetic powder clutch and a rope storage wheel connected in sequence, as well as an output-side encoder and a tension sensor, characterized in that, The controller updates the rope storage status based on the initial rope storage volume and the signed position increment. It queries the rope storage length and effective loading lever arm using the same channel's updated status snapshot and calibration version, and calculates the rope take-up and release length. Different target tensions are set for channels working simultaneously, and the loading is controlled separately based on the effective loading lever arm and the measured tension. The timeliness, acquisition time and status, and version correspondence of each necessary channel sample are checked. After all checks are passed, the associated results are submitted. The error between the measured tension and the target tension of a single channel and the error between the channel's measured tension difference and the target tension difference are verified. After the error, data correspondence, and control quantity constraints continuously meet the set conditions and the observation time is reached, the load condition mark is associated with the length, rope storage status, and calibration version and recorded. Non-compliance and transition records are retained.

2. The multi-channel rope reeling and length counting device according to claim 1, characterized in that, The rope storage wheel has a rope groove that constrains the axial parallel winding of the rope, so that the rope is stored in a single-row radial overlapping manner; the controller uses the initial rope storage amount determined after installation as a reference and uses the same rope storage state reference during the rope winding and unwinding process; the encoder is coaxially connected to the rope storage wheel, or connected through a transmission mechanism with a known transmission ratio, and does not use the rotation amount on the input side of the magnetic powder clutch to replace the actual rotation amount of the rope storage wheel.

3. The multi-channel rope reeling and length counting device according to claim 1, characterized in that, The controller unwinds the difference between the counts at adjacent positions based on the encoder counting modulus and forms a signed cumulative count according to the installation direction; the product of the maximum allowable sampling interval and the maximum rate of change of count is less than half of the counting modulus; if the sampling timeout, the displacement increment exceeds the allowable range, or there is a half-range ambiguity in the count difference, the sample is rejected and the valid rope storage status is not updated based on it.

4. The multi-channel rope reeling and length counting device according to claim 1, characterized in that, The relationship between the rope length and the effective loading lever arm is composed of calibration nodes arranged according to the number of rope turns, with a common channel identifier and calibration version. The rope length and effective loading lever arm are obtained by interpolation of the number of rope turns in the same snapshot. The two relationships are established separately with the actuator calibration relationship from the magnetic powder control quantity to the transmitted torque. Extrapolation is rejected when the effective calibration range is exceeded, and the calibration range abnormality is recorded.

5. The multi-channel rope reeling and length counting device according to claim 1, characterized in that, The controller generates a feedforward torque based on the product of the target tension and the effective loading arm. It then synthesizes this torque with the proportional-integral feedback correction of the tension error. This is converted into a magnetic particle control quantity through the corresponding actuator calibration relationship, and amplitude and rate of change limits are implemented. If using the current integral candidate value would cause the control quantity to exceed the amplitude or rate of change limits, the previous integral value is retained and the control quantity is recalculated. Each channel saves its integral state and the previous control quantity.

6. The multi-channel rope reeling and length counting device according to claim 1, characterized in that, The controller sets the allowable tension error for each channel and the allowable load difference error for each channel pair. The allowable load difference error of at least one selected channel pair is less than the sum of the allowable tension errors of the two channels in the channel pair. The corresponding sample is allowed to enter the load condition observation window only when both the single channel error verification and the channel pair load difference error verification are passed.

7. The multi-channel rope reeling and unloading differential loading and length counting device according to claim 1, characterized in that, The data correspondence includes channel identifier, status update sequence number, acquisition time, load formula version and calibration version; the controller terminates the original observation window when the formula changes, the calibration version changes, the release direction is detected to change or an invalid sample appears, and the observation window is re-established for consecutive samples that meet the conditions. When a stationary sample is encountered, the direction judgment retains the previous non-zero motion direction.

8. A multi-channel rope take-up and release differential loading and length counting method, applied to at least two rope take-up and release channels having a drive unit, a magnetic powder clutch, a rope storage wheel, an output-side encoder, and a tension sensor, characterized in that, include: The system collects the position of the rope storage wheel on the output side of the magnetic powder clutch and the measured tension of each channel. The rope storage status is updated based on the initial rope storage amount and the increment of the signed position. The rope storage status and calibration version for the same channel and the same update are fixed. The rope length relationship and effective loading arm relationship are queried separately. The rope length for take-up and release is calculated based on the difference between the initial and current rope lengths. Different target tensions are set for at least two simultaneously operating channels. A loading reference is formed based on each target tension and the effective loading arm, and the magnetic powder clutch control quantity is adjusted according to the measured tension. Timeliness, time deviation, and status and version correspondence are checked for each necessary channel sample. After all necessary channels pass, the results of this correlation calculation are submitted. The error of the measured tension relative to the target tension for each channel, as well as the error of the measured tension difference relative to the target tension difference for the selected channel, are verified. When the errors, data correspondence, and control quantity constraints continuously meet the set conditions and the observation time is reached, the corresponding sampling window is marked as meeting the load conditions. This mark is recorded together with the corresponding rope length for take-up and release, rope storage status, and calibration version, and records of non-compliance and transitions are retained.

9. A method for differential loading and length counting of multi-channel rope reeling and unloading according to claim 8, characterized in that, The candidate position, length, effective lever arm, and load calculation results are temporarily stored and then checked accordingly. If the check fails, the new state of some channels is not used to form the association result of the checked channels with the old state of other channels. If the actual rope storage state cannot be determined due to counting ambiguity or loss of continuity, the rope storage state is set to the pending confirmation state. After reconfirming the correspondence between the encoder count and the rope storage amount, the effective state is restored and updated.

10. A method for differential loading and length counting of multi-channel rope reeling and unloading according to claim 8, characterized in that, For each sampling, the original location, sampling time, target tension, measured tension, rope status, length counting result, status update number, formula version, calibration version, control quantity constraint flag, and load condition mark or invalid reason are saved. When the formula, calibration, direction, or validity conditions change, the original window is terminated, and the start and end numbers of the window and the corresponding length range are saved. The terminated window and the reconstructed window are not merged into a continuous matching window.

Images