A method for measuring belt deviation based on intelligent sensors

Abstract

The application relates to the technical field of industrial conveying equipment state measurement, and discloses a belt offset measurement method based on an intelligent sensor. Multiple measurement positions and unified longitudinal and lateral reference are established in a straight conveying section, path distance, sensor effective range, resolution and reference edge distance are obtained, longitudinal velocity and left and right edge distance are uniformly sampled, cumulative travel and a belt entity reference section are established, a propagation arrival time is calculated according to a target cumulative travel, an upper limit arrival sampling index and a distance position ratio, multiple position propagation alignment of the same entity section is realized, offset across positions is formed through effective interval screening and resolution weighting, offset span, cumulative offset change and lateral net change are calculated, and finally, a measurement record is generated, so that the influence of multiple position object misplacement, abnormal data and measurement capacity difference is reduced, and the comparability, traceability and engineering interpretability of the offset result are improved.

Description

Technical Field

[0001] This invention relates to the field of industrial conveyor equipment condition measurement technology, specifically a method for measuring belt offset based on intelligent sensors. Background Technology

[0002] Belt conveyors are widely used in material handling in mines, ports, metallurgy, logistics, and production lines. During continuous operation, belts are prone to lateral shift due to factors such as uneven material distribution, idler installation deviations, changes in tension, differences in local resistance, and long-term wear. If this lateral shift continues to increase, it can cause the belt edge to contact the frame, material spillage, abnormal edge wear, and affect the stable operation of the conveying system. Therefore, continuous and accurate measurement of the lateral shift of the running belt is a crucial aspect of belt conveyor condition monitoring.

[0003] Existing belt misalignment detection methods typically employ measurement units such as distance sensors, displacement sensors, and vision inspection devices at one or more fixed locations along the conveyor path. The misalignment is determined by detecting changes in the position of the belt's left and right edges relative to a fixed reference. While single-location detection can capture localized lateral misalignment, it struggles to reflect the overall misalignment of the same belt segment as it propagates along the conveying direction. Multi-location detection usually acquires data from different measurement locations simultaneously at a unified sampling time, then compares, statistically analyzes, or fuses the results to determine the overall degree of belt misalignment. Some solutions also categorize detection results into normal, warning, and abnormal states by setting a fixed misalignment threshold. However, belts are solid materials undergoing continuous longitudinal movement. When multiple measurement locations are sampled simultaneously, each location corresponds to a different cross-sectional position along the belt's length. Directly comparing lateral misalignment data acquired at the same sampling time can easily lead to misinterpreting edge differences between different belt locations as lateral changes in the same cross-section during propagation, resulting in measurement object mismatch in cross-location misalignment analysis. Especially when belt speed changes, there are brief pauses followed by restarts, and the longitudinal spacing between measurement positions is large, it becomes increasingly difficult to accurately establish the correspondence between the same cross-section at different measurement positions simply based on a fixed time difference or synchronous sampling relationship. Furthermore, in existing multi-position measurements, the sensor installation reference, left-right measurement distance, and resolution may differ at each measurement position. Directly using the original left-right edge distances or simple averaging results is easily affected by differences in fixed installation and measurement capabilities. Simultaneously, providing only a single offset result makes it difficult to simultaneously reflect the maximum offset range, cumulative lateral change, and net change between the first and last positions of the same belt cross-section during its propagation from upstream to downstream.

[0004] Therefore, this case aims to propose a belt offset measurement method based on intelligent sensors. The method first establishes multiple intelligent sensor measurement positions along the conveyor path and completes the unified calibration of the measurement space using longitudinal baselines, transverse measurement baselines, baseline edge distances, effective measurement closed intervals, and sensor resolution. Then, the longitudinal conveying speed of the belt and the left and right edge distances at each position are uniformly sampled. The time-sampled data is converted into a propagation basis for tracking the actual position of the belt by accumulating the travel distance. A belt physical reference cross-section is established based on the belt cross-sectional position at a certain sampling time at the first measurement position. Subsequently, instead of directly comparing the offsets obtained at the same time from different measurement positions, the arrival time of the same physical reference cross-section propagating to different measurement positions is calculated. The transverse offsets at each measurement position are propagated and aligned, and then combined with the sensor resolution to form the contribution share across measurement positions, resulting in a comprehensive offset representing the same physical cross-section. Simultaneously, the cross-position offset span, cumulative offset change, and net transverse change are extracted, ultimately forming a structured measurement record arranged according to the effective reference cross-section sampling index. Summary of the Invention

[0005] This invention provides a method for measuring belt offset based on intelligent sensors, which helps to solve the problems mentioned in the background art.

[0006] This invention provides the following technical solution: a method for measuring belt offset based on a smart sensor, comprising:

[0007] In the straight conveying measurement section, establish and number multiple intelligent sensor measurement positions, establish longitudinal and transverse measurement baselines for the conveying path, obtain the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and baseline edge distances, form effective measurement closed intervals, and calculate the baseline distance difference;

[0008] Uniform sampling is performed on the longitudinal conveying speed measurement, left edge distance, and right edge distance. The constant equivalent longitudinal conveying speed and cumulative travel are calculated. The belt physical reference section, reference section sampling index, and reference section sampling time are established. The effective reference section sampling index is then selected for propagation.

[0009] Calculate the target cumulative journey based on the cumulative journey and path distance, retrieve the upper bound sampling index, and calculate the distance-location ratio, propagation arrival time, and propagation time difference;

[0010] Read the left edge distance and right edge distance and verify the corresponding valid measurement closed interval. Record the valid reference section sampling index of the propagation that has passed the verification as the valid reference section sampling index of the edge distance. Calculate the belt lateral offset and the propagation alignment lateral offset.

[0011] The combined resolution is calculated based on the resolution of the left and right distance sensors, and then normalized by the inverse of the combined resolution to form the representative offset contribution share across the measurement positions.

[0012] The lateral offset is aligned by weighted cumulative propagation based on the offset contribution share across measurement locations to form the representative offset across measurement locations.

[0013] Calculate the cross-position offset span, cross-position cumulative offset change, and net lateral change based on the propagation alignment lateral offset;

[0014] The propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change are summarized to generate a belt offset measurement record, which is then output in ascending order of edge distance effective reference section sampling index.

[0015] Optionally, the step of establishing and numbering multiple intelligent sensor measurement positions in the straight conveying measurement section, establishing a longitudinal baseline and a transverse measurement baseline for the conveying path, obtaining the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and the baseline edge distance, forming an effective measurement closed interval, and calculating the baseline distance difference, specifically includes:

[0016] In the linear conveyor measurement section, at least two intelligent sensor measurement positions are established sequentially along the belt running direction. The measurement positions are numbered according to their arrangement order. The total number of intelligent sensor measurement positions included in the current measurement is counted, and a longitudinal baseline of the conveyor path is established that runs through all measurement positions.

[0017] Set the first measurement position as the longitudinal distance reference, read the path distance of each measurement position relative to the first measurement position in sequence along the longitudinal reference line of the conveying path, record the path distance of the first measurement position as zero, and record the path distance of subsequent measurement positions in ascending order of measurement position number.

[0018] Facing the direction of belt running, the right side of the belt is defined as the positive direction and the left side as the negative direction. At each measurement position, a transverse measurement baseline perpendicular to the longitudinal baseline of the conveying path is established. On each transverse measurement baseline, a left fixed measurement baseline and a right fixed measurement baseline are established relative to the conveying device. The distance from the left fixed measurement baseline to the left edge of the belt and the distance from the right edge of the belt to the right fixed measurement baseline are read along the corresponding transverse measurement baseline.

[0019] Read the effective lower limit and effective upper limit of the left and right distance sensors at each measurement position respectively. The effective lower limit and effective upper limit of the same distance sensor constitute the effective measurement closed interval of the corresponding distance sensor. Compare the effective lower limit and effective upper limit of the same distance sensor. The effective measurement boundary condition is that the effective lower limit is a non-negative value and the effective lower limit is less than the corresponding effective upper limit.

[0020] When the left or right distance sensor at any measurement position does not meet the valid measurement boundary conditions, the current measurement is determined to be invalid and the generation of the belt offset measurement record corresponding to the current measurement is stopped. When the left and right distance sensors at all measurement positions meet the valid measurement boundary conditions, under the centered calibration state where the longitudinal center line of the belt coincides with the longitudinal reference line of the conveying path, the left reference edge distance and right reference edge distance of each measurement position are read and compared with the effective lower limit and effective upper limit of the corresponding distance sensor respectively.

[0021] When any left or right reference edge distance is outside the effective measurement closed interval of the corresponding distance sensor, the current measurement is deemed invalid and the generation of belt offset measurement records is stopped. When all left and right reference edge distances are within the effective measurement closed interval of their respective corresponding distance sensors, the left reference edge distance is subtracted from the right reference edge distance for each measurement position to form the reference distance difference of the current measurement position.

[0022] Read the resolution of the left and right distance sensors at each measurement position respectively, compare the resolution of the left and right distance sensors at each measurement position with zero respectively, and check whether the resolution of the left and right distance sensors at each measurement position is positive.

[0023] When the resolution of the left and right distance sensors at all measurement positions is positive, the resolution of the left and right distance sensors at each measurement position is written into the cross-measurement position offset contribution share calculation data. When the resolution of the left or right distance sensor at any measurement position is not positive, the current measurement is determined to be invalid and the generation of belt offset measurement records is stopped.

[0024] Optionally, the step of uniformly sampling the measured longitudinal conveying speed, left edge distance, and right edge distance, calculating the constant equivalent longitudinal conveying speed and cumulative travel, establishing the belt physical reference section, reference section sampling index, and reference section sampling time, and filtering the propagation of valid reference section sampling indexes specifically includes:

[0025] Read the unified sampling period, the start sampling time of the current sampling batch, and the sampling quantity of the current sampling batch. Check whether the unified sampling period is a positive real number and check whether the sampling quantity of the current sampling batch is not less than two. Use the fact that the unified sampling period is a positive real number and the sampling quantity of the current sampling batch is not less than two as the sampling configuration conditions.

[0026] When the current sampling batch does not meet the sampling configuration conditions, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When the current sampling batch meets the sampling configuration conditions, the unified sampling time is generated sequentially according to the unified sampling period starting from the initial sampling time of the current sampling batch. The discrete sampling index is established from zero according to the time sequence. The unified longitudinal conveying speed measurement value of the physical material on the belt surface, as well as the left edge distance and right edge distance of all measurement positions are uniformly written into the corresponding unified sampling time.

[0027] For each uniform sampling moment, a straight conveying measurement section with the same longitudinal conveying speed is used at each cross-sectional position of the solid material on the belt surface. The uniform longitudinal conveying speed measurement value obtained directly from the solid material on the belt surface is read. The direction of the uniform longitudinal conveying speed measurement value is specified according to the increasing direction of the measurement position number. Each uniform longitudinal conveying speed measurement value is checked one by one to see if it is non-negative.

[0028] When there is a negative uniform longitudinal conveying speed measurement value, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When all uniform longitudinal conveying speed measurements are non-negative, for every two adjacent uniform sampling times, the time interval between the two adjacent uniform sampling times is recorded as the corresponding sampling interval. The uniform longitudinal conveying speed measurement value of the previous uniform sampling time is added to the uniform longitudinal conveying speed measurement value of the next uniform sampling time, and then the result is divided by two to calculate the constant equivalent longitudinal conveying speed within the corresponding sampling interval.

[0029] The cumulative travel distance at the start sampling time of the current sampling batch is recorded as zero. For each sampling interval after the start sampling time of the current sampling batch, the corresponding constant equivalent longitudinal transmission speed is multiplied by the unified sampling period to calculate the longitudinal propagation distance of the corresponding sampling interval. Then, the longitudinal propagation distance of the corresponding sampling interval is added to the cumulative travel distance of the previous unified sampling time to calculate the cumulative travel distance corresponding to each unified sampling time, forming a cumulative travel distance sequence arranged according to the unified sampling time. The propagation arrival time within the corresponding sampling interval is processed according to the constant equivalent longitudinal transmission speed of the corresponding sampling interval.

[0030] For each unified sampling time within the current sampling batch, the discrete sampling index corresponding to the unified sampling time is recorded as the reference section sampling index, the cross-sectional position of the belt surface material at the first measurement position corresponding to the unified sampling time is recorded as the belt entity reference section of the corresponding reference section sampling index, the unified sampling time is recorded as the reference section sampling time, and the cumulative travel corresponding to the reference section sampling time is recorded.

[0031] For each reference section sampling index corresponding to the belt entity reference section before the last unified sampling time, compare the cumulative travel of the previous unified sampling time with the cumulative travel of the current reference section sampling time. Only retain the reference section sampling index corresponding to the cumulative travel of the previous unified sampling time being greater than the cumulative travel of the current reference section sampling time. Then, subtract the cumulative travel of the corresponding reference section sampling time from the cumulative travel of the last unified sampling time of the current sampling batch to calculate the remaining cumulative travel. Compare the remaining cumulative travel with the path distance of the last measurement position relative to the first measurement position. Only retain the reference section sampling index with the remaining cumulative travel not less than the path distance of the last measurement position relative to the first measurement position. Record the retained reference section sampling index as the propagation effective reference section sampling index, and form the propagation effective reference section sampling index set from the propagation effective reference section sampling index.

[0032] When the set of effective reference section sampling indexes for propagation is empty, the generation of belt offset measurement records for the current sampling batch is stopped. When the set of effective reference section sampling indexes for propagation is not empty, the propagation arrival time, propagation time difference, and propagation alignment lateral offset of each measurement position are calculated one by one according to the effective reference section sampling indexes for propagation.

[0033] Optionally, the step of calculating the target cumulative journey based on the cumulative journey and path distance, retrieving the upper bound sampling index, and calculating the distance-position ratio, propagation arrival time, and propagation time difference specifically includes:

[0034] For each belt entity reference section corresponding to the sampling index of the effective reference section, the cumulative travel time of the reference section corresponding to the current belt entity reference section is read. For each measurement position from the second measurement position to the last measurement position, the cumulative travel time of the reference section corresponding to the current belt entity reference section at the sampling time is added to the path distance of the current measurement position relative to the first measurement position to calculate the target cumulative travel time when the current belt entity reference section propagates to the current measurement position.

[0035] Read the sampling time of the reference section corresponding to the current reference section of the belt entity, record the sampling time of the reference section as the propagation arrival time of the first measurement position, and record the propagation time difference of the current reference section of the belt entity at the first measurement position as zero;

[0036] For each measurement position from the second measurement position to the last measurement position, starting from the cumulative travel sequence after the sampling time of the corresponding reference section, each cumulative travel is compared with the corresponding target cumulative travel one by one according to the sampling time sequence, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is retrieved, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is recorded as the upper bound sampling index.

[0037] Read the cumulative travel distance corresponding to the upper bound sampling index and the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the remaining propagation distance of the target cumulative travel distance relative to the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the cumulative travel distance increment between the target and the previous discrete sampling index. Divide the remaining propagation distance of the target cumulative travel distance relative to the previous discrete sampling index by the cumulative travel distance increment between the target and the previous discrete sampling index to calculate the distance-position ratio.

[0038] Read the unified sampling time corresponding to the upper bound sampling index and the unified sampling time corresponding to the previous discrete sampling index. Subtract the unified sampling time corresponding to the previous discrete sampling index from the unified sampling time corresponding to the upper bound sampling index to calculate the sampling time length between the upper bound sampling index and the previous discrete sampling index. Multiply the sampling time length between the upper bound sampling index and the previous discrete sampling index by the distance-position ratio to calculate the time increment. Add the time increment to the unified sampling time corresponding to the previous discrete sampling index to calculate the propagation arrival time of the belt entity reference section to the corresponding measurement position.

[0039] For each measurement position from the second measurement position to the last measurement position, the propagation time difference from the first measurement position to the corresponding measurement position is calculated by subtracting the sampling time of the reference section corresponding to the current belt entity reference section from the corresponding propagation arrival time.

[0040] Optionally, the step of reading the left edge distance and right edge distance and verifying the corresponding valid measurement closed interval, recording the valid propagation reference section sampling index that has passed the verification as the edge distance valid reference section sampling index, and calculating the belt lateral offset and propagation alignment lateral offset specifically includes:

[0041] For each belt entity reference section corresponding to the effective reference section sampling index, read the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section, and record the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section as the left edge distance and right edge distance of the first measurement position actually participating in the lateral offset reconstruction.

[0042] For each measurement position from the second measurement position to the last measurement position, the corresponding distance-position ratio is read. When the distance-position ratio is equal to one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index are read. When the distance-position ratio is greater than zero and less than one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index and the left edge distance and right edge distance at the unified sampling time corresponding to the previous discrete sampling index are read respectively. The left edge distance and right edge distance read according to the distance-position ratio are recorded as the left edge distance and right edge distance of the corresponding measurement position that actually participate in the lateral offset reconstruction.

[0043] For each left edge distance that actually participates in the lateral offset reconstruction, the current left edge distance is compared with the effective lower limit and effective upper limit of the corresponding left distance sensor. For each right edge distance that actually participates in the lateral offset reconstruction, the current right edge distance is compared with the effective lower limit and effective upper limit of the corresponding right distance sensor.

[0044] When all measured positions are within the effective measurement closed interval of their respective distance sensors, the left and right edge distances that actually participate in the lateral offset reconstruction are recorded as the edge distance effective reference section sampling index. When any left or right edge distance that actually participates in the lateral offset reconstruction is outside the effective measurement closed interval of the corresponding distance sensor, the measurement result of the current propagation effective reference section sampling index is determined to be invalid edge distance. The calculation of belt lateral offset, propagation alignment lateral offset, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, and lateral net change corresponding to the current propagation effective reference section sampling index is stopped, and the generation of the corresponding belt offset measurement record is stopped.

[0045] For each group of actual left and right edge distances participating in the lateral offset reconstruction corresponding to the effective reference section sampling index of edge distance, the left edge distance is successively subtracted from the right edge distance and the reference distance difference formed by the corresponding measurement position in the centering calibration state. The difference is then divided by two to calculate the belt lateral offset at the corresponding measurement position at the corresponding unified sampling time. Belt lateral offsets greater than zero are recorded as belt lateral offset to the right, belt lateral offsets less than zero are recorded as belt lateral offset to the left, and belt lateral offsets equal to zero are recorded as no lateral offset formed relative to the centering calibration state at the corresponding measurement position and unified sampling time.

[0046] The left edge distance and right edge distance corresponding to the sampling time of the reference section are read at the first measurement position. The belt lateral offset corresponding to the sampling time of the reference section is calculated by subtracting the right edge distance and the difference between the reference distance at the first measurement position from the left edge distance and then dividing by two.

[0047] For each measurement position from the second to the last measurement position, when the distance-position ratio is equal to one, the belt lateral offset at the unified sampling time corresponding to the upper bound sampling index is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index. When the distance-position ratio is greater than zero and less than one, the belt lateral offset corresponding to each of the two adjacent unified sampling times is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the previous discrete sampling index of the upper bound sampling index, as well as the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index.

[0048] For the first measurement position, the lateral offset of the belt corresponding to the sampling time of the reference section is directly recorded as the propagation alignment lateral offset of the current belt entity reference section at the first measurement position. For the second to the last measurement position, when the distance-position ratio is equal to one, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is directly recorded as the propagation alignment lateral offset of the corresponding measurement position. When the distance-position ratio is greater than zero and less than one, the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index is calculated by subtracting the distance-position ratio from one. The lateral offset of the belt at the unified sampling time corresponding to the previous discrete sampling index is multiplied by the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index. Then, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is multiplied by the distance-position ratio. The two products are added together to calculate the propagation alignment lateral offset of the current belt entity reference section at the corresponding measurement position.

[0049] Optionally, the step of calculating the combined resolution based on the resolutions of the left and right distance sensors, and normalizing it by the inverse of the combined resolution to form the representative offset contribution share across measurement positions, specifically includes:

[0050] For all measurement locations, the resolutions of the left and right distance sensors, which serve as the data for calculating the contribution share of offset across measurement locations, are read respectively.

[0051] For each measurement location, the resolution of the left distance sensor and the resolution of the right distance sensor at the current measurement location are added together to calculate the combined resolution of the left and right distance sensors at the current measurement location.

[0052] For each measurement location, the reciprocal of the combined resolution of the left and right distance sensors at the current measurement location is taken to calculate the resolution weight base value corresponding to the current measurement location;

[0053] The resolution weight base values ​​corresponding to all measurement locations are added together sequentially to calculate the total resolution weight base values ​​of all measurement locations;

[0054] For each measurement location, the resolution weight base value corresponding to the current measurement location is divided by the sum of the resolution weight base values ​​of all measurement locations to calculate the cross-measurement location representative offset contribution share of the current measurement location.

[0055] Optionally, the step of weighting and accumulating the propagation alignment of the lateral offset according to the representative offset contribution share across measurement locations to form a representative offset across measurement locations specifically includes:

[0056] For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, read the propagation alignment lateral offset of all measurement positions and the cross-measurement position representative offset contribution share of each measurement position in ascending order of measurement position number;

[0057] For each measurement location, the propagation alignment lateral offset of the current measurement location is multiplied by the cross-measurement location representative offset contribution share corresponding to the current measurement location to calculate the representative offset component corresponding to the current measurement location.

[0058] The representative offset components corresponding to all measurement positions are added together in sequence to calculate the representative offset across the measurement positions of the current belt entity reference section. The representative offset across the measurement positions that is greater than zero is recorded as the result of being biased to the right side of the belt, and the representative offset across the measurement positions that is less than zero is recorded as the result of being biased to the left side of the belt.

[0059] Optionally, the calculation of the cross-position offset span, the cumulative cross-position offset change, and the net lateral change based on the propagation alignment lateral offset specifically includes:

[0060] For each belt entity reference section corresponding to the effective reference section sampling index at each edge distance, read the propagation alignment lateral offset of all measurement positions, filter out the maximum and minimum propagation alignment lateral offsets from all measurement positions, subtract the minimum propagation alignment lateral offset from the maximum propagation alignment lateral offset, and calculate the cross-position offset span of the current belt entity reference section.

[0061] According to the ascending order of the measurement position numbers, for each group of adjacent measurement positions, the propagation alignment lateral offset of the next measurement position is subtracted from the propagation alignment lateral offset of the previous measurement position. The absolute value of the difference is taken, and then the absolute values ​​corresponding to all adjacent measurement positions are added together to calculate the cumulative offset change of the current belt entity reference section across the position.

[0062] Subtract the lateral offset of the propagation alignment from the first measurement position from the lateral offset of the propagation alignment at the last measurement position to calculate the net lateral change of the current belt entity reference section from the first measurement position to the last measurement position. Record the net lateral change that is greater than zero as the net change of the current belt entity reference section to the right lateral side of the belt after propagation from the first measurement position to the last measurement position, and record the net lateral change that is less than zero as the net change of the current belt entity reference section to the left lateral side of the belt after propagation from the first measurement position to the last measurement position.

[0063] Optionally, the step of summarizing the propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change to generate a belt offset measurement record, and outputting it in ascending order of edge distance effective reference section sampling index, specifically includes:

[0064] For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, the propagation alignment lateral offset of each measurement position is read in ascending order of measurement position number, and all propagation alignment lateral offsets are arranged in ascending order of measurement position number to generate multi-position propagation alignment offset ordered data corresponding to the current belt entity reference section.

[0065] According to the ascending order of the measurement position numbers, the propagation time difference of the current belt physical reference section from the first measurement position to each measurement position is read sequentially, and all propagation time differences are arranged in ascending order of the measurement position numbers to generate ordered data of the propagation time difference corresponding to the current belt physical reference section.

[0066] The system sequentially reads the current edge distance effective reference section sampling index, multi-position propagation alignment offset ordered data, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, lateral net change, and propagation time difference ordered data. It then combines these data according to the correspondence of the same belt entity reference section to generate the belt offset measurement record corresponding to the current edge distance effective reference section sampling index.

[0067] The belt offset measurement records are arranged and output in ascending order of the effective reference section sampling index based on the edge distance.

[0068] The present invention has the following beneficial effects:

[0069] 1. By constructing a lateral offset measurement benchmark system that maintains a consistent meaning across multiple locations, including the longitudinal benchmark of the conveyor path, the lateral measurement benchmarks at each measurement position, the left and right fixed measurement benchmarks, the left and right edge distances, the effective measurement upper and lower limits, the effective measurement closed interval, the benchmark edge distance under centered calibration, and the benchmark distance difference, a lateral offset measurement benchmark system is established. The establishment of the benchmark distance difference allows subsequent lateral offset calculations to operate without relying on perfect symmetry of the left and right installation distances. Instead, it allows for actual installation differences at each measurement position, as long as benchmark compensation is completed under centered calibration. This reduces the systematic impact of sensor mechanical installation errors, imperfect symmetry of the left and right fixed benchmarks, and differences in structural dimensions at each location on the offset results. Furthermore, by pre-constraining the effective measurement upper and lower limits and the effective measurement closed interval, out-of-range data can be excluded before calculating the offset, preventing abnormal edge distances from being incorrectly interpreted as actual belt offset.

[0070] 2. This scheme adopts a unified sampling sequence for longitudinal conveying speed and left and right edge distances. It then forms an equivalent longitudinal conveying speed within the sampling interval based on speed measurements at adjacent sampling times, and accumulates this to obtain the actual belt propagation distance. The cross-sectional position of the belt material at a specific sampling time at the first measurement position is recorded as the belt's reference cross-section, ensuring that all subsequent measurements revolve around this entity. This transforms multi-position measurements from spatial comparisons at the same time into propagation comparisons of the same entity cross-section, fundamentally solving the data object misalignment problem caused by belt movement. Furthermore, this scheme filters the sampling index of the effective reference cross-section for propagation, retaining only the entity cross-section that can propagate to the final measurement position within the current sampling batch, avoiding incomplete data or partial coverage of measurement positions. This design, which first filters for propagation reachability, helps ensure the integrity of the data structure for subsequent cross-position results.

[0071] 3. The scheme combines the cumulative travel corresponding to the sampling time of the reference section with the path distance of the measurement position relative to the first measurement position to form the target cumulative travel. Then, it retrieves the arrival upper bound sampling index from the cumulative travel sequence for the first time reaching or exceeding the target cumulative travel. Using the cumulative travel relationship between the arrival upper bound sampling index and the previous discrete sampling index, it forms a distance-position ratio, further calculating the propagation arrival time and propagation time difference. The actual operating speed of the conveyor belt may vary. If the propagation delay is simply calculated by dividing the distance between measurement positions by a fixed speed, speed fluctuations will accumulate into significant time misalignment. This scheme directly forms the cumulative travel based on the actual longitudinal speed collected and further estimates the specific arrival position between adjacent sampling points, thus adapting to operating conditions where speed changes over time. Compared to schemes using fixed delays or whole-point alignment, this method refines the propagation arrival time to the interval between two adjacent sampling times, rather than forcing the arrival event to fall on a single discrete sampling point, thereby reducing the time quantization error caused by the sampling period itself. For scenarios with high-speed conveying, limited sampling periods, and large measurement position intervals, this propagation time reconstruction mechanism has very direct engineering value.

[0072] 4. The scheme first compares the actual left and right edge distances involved in the lateral offset reconstruction with the effective measurement closed intervals of the corresponding distance sensors. Only the sampling index of the effective reference section that meets the effective conditions at all measurement positions is further recorded as the edge distance effective reference section sampling index. This design ensures that subsequent cross-position calculations are based on the premise that "the same physical cross section has effective distance measurement data at all participating positions." Subsequently, the scheme calculates the belt lateral offset at each measurement position using the reference distance difference. For cases where the arrival time is between adjacent unified sampling times, the offset of adjacent sampling times is propagated and aligned for reconstruction by using the distance-position ratio. This avoids contaminating the cross-position results with invalid edge distances; avoids time deviation caused by simply taking the nearest sampling point; and ensures that the outputs at different positions represent the true offset state corresponding to the same physical reference section. Compared with traditional methods that directly read the current offset of a sensor or use the nearest sample value to replace the actual propagation time, this scheme improves the physical consistency and temporal correspondence accuracy of cross-position data.

[0073] 5. This scheme does not treat multiple measurement locations as equivalent nodes with identical measurement quality. Instead, it utilizes the resolution of the sensors at the left and right distances of each measurement location to form a combined resolution. Furthermore, it uses the inverse of the combined resolution, after normalization, to determine the cross-measurement location representative offset contribution share. This design embodies the idea that "differences in measurement capabilities should be incorporated into the multi-location fusion process." When different locations use different sensor models, have different installation conditions, detection distances, or actual resolution capabilities, a simple arithmetic average would assume that each location has the same measurement reliability, potentially allowing locations with weaker resolution capabilities to have the same impact on the final integrated result as locations with higher resolution capabilities. This scheme constructs contribution shares through resolution, allowing measurement locations with better resolution capabilities to contribute more reasonably to the cross-location representative offset, while normalization ensures that the contribution relationship of all measurement locations has a uniform scale. Compared to traditional averaging, majority voting, or manually assigned fixed weights, the advantage of this scheme is that the contribution relationship directly originates from the quantifiable measurement attributes of the sensors themselves, rather than being set by human experience. Therefore, the source of the weights is more objective and easier to interpret, and it is also convenient to automatically re-form the corresponding contribution relationship after different equipment configurations or sensor replacements.

[0074] 6. After propagation alignment is completed, this scheme further weights and accumulates the lateral offsets of each measurement location according to their contribution share across measurement locations, thus forming a "representative offset across measurement locations for the same belt body reference section." The fundamental difference between this process and ordinary multi-point averaging is that what is fused is not the original offset data obtained at the same time from each measurement location, but rather the offset states of multiple locations that have been reconstructed after propagation and now belong to the same physical reference section. Therefore, the fused object has physical uniformity. Simultaneously, the contribution of each measurement location is determined by the sensor resolution capability of the previous step. First, it can compress multiple local measurement results into a single offset index with overall representativeness, facilitating trend judgment, status display, and alarm strategy formulation by the subsequent monitoring system. Second, this representative value is not a simple average, but is based on propagation alignment and measurement capability weighting, thus better reflecting the overall offset tendency of the same belt section throughout the entire measurement area. Compared to single-point sensors that only reflect local conditions, it reduces the dominance of random fluctuations at a single measurement point on the results; compared to multi-point averaging without propagation alignment, it avoids mixing the offset states of different physical cross sections into the same value, thus having stronger physical interpretability.

[0075] 7. This scheme goes beyond simply outputting a single representative offset across measurement positions. Instead, it extracts three complementary variation characteristics from the propagation and alignment lateral offset of the same belt's reference cross-section across multiple measurement positions: cross-position offset span, cumulative cross-position offset change, and net lateral change. The cross-position offset span reflects the overall variation between the maximum and minimum offsets of the same cross-section within the entire measurement section. The cumulative cross-position offset change, through the accumulation of the absolute values ​​of the offset differences between adjacent measurement positions, reflects the number and extent of lateral changes during propagation. The net lateral change directly indicates the net offset to the left or right after propagation from the first measurement position to the last measurement position. These three indicators can distinguish between seemingly identical "final offsets" but completely different propagation behaviors. For example, one belt segment might initially deflect to the right and then swing back to the left, resulting in a small net change but a large cumulative offset change; another belt segment might consistently drift slowly to one side, with a relatively stable cumulative change but a significant net change. Traditional methods that only consider single-point offset or maximum offset are insufficient to distinguish these states. This solution, however, can describe the offset propagation pattern in greater depth, providing a richer data foundation for identifying different problems such as continuous deviation, reciprocating oscillation, and local abrupt changes.

[0076] 8. This scheme organizes the aforementioned propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change according to the correspondence of the same belt entity reference section. It outputs the data in ascending order, using the edge distance effective reference section sampling index as the main recording line. Its value lies not merely in "saving data," but in establishing a structured result organization method oriented towards entity section tracking. This ensures that each measurement record corresponds to a specific belt entity section that has passed propagation reachability and edge distance validity verification. Subsequent analysis, including trend analysis, anomaly backtracking, equipment diagnosis, and alarm strategy judgment, can trace the propagation state of this belt segment across multiple measurement positions along the entity reference section sampling index, rather than simply obtaining isolated instantaneous sensor readings. Compared to traditional systems that save multiple sensor raw values ​​separately by timestamp, the data output by this scheme has already undergone spatiotemporal alignment, validity filtering, and feature extraction. This results in a higher data level, a lower subsequent analysis burden, and easier establishment of continuous offset evolution records. Especially when it is necessary to analyze the process of offset formation of a certain section of belt from upstream to downstream, this method of organizing data according to the physical cross section is more meaningful for engineering diagnosis than simply organizing data according to equipment channel or sampling time. Attached Figure Description

[0077] Figure 1 This is a schematic diagram showing the arrangement of the linear conveying measurement section and the intelligent sensor measurement positions according to the present invention.

[0078] Figure 2 This is a schematic diagram illustrating the unified sampling and establishment of the reference cross section of the belt entity in this invention.

[0079] Figure 3 This is a schematic diagram illustrating the calculation of the propagation arrival time and the lateral offset of propagation alignment in this invention.

[0080] Figure 4 This is a schematic diagram illustrating the generation of measurement records for representative offsets and belt offsets across measurement positions according to the present invention. Detailed Implementation

[0081] 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.

[0082] An embodiment of a belt offset measurement method based on smart sensors includes:

[0083] In the straight conveying measurement section, establish and number multiple intelligent sensor measurement positions, establish longitudinal and transverse measurement baselines for the conveying path, obtain the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and baseline edge distances, form effective measurement closed intervals, and calculate the baseline distance difference;

[0084] Uniform sampling is performed on the longitudinal conveying speed measurement, left edge distance, and right edge distance. The constant equivalent longitudinal conveying speed and cumulative travel are calculated. The belt physical reference section, reference section sampling index, and reference section sampling time are established. The effective reference section sampling index is then selected for propagation.

[0085] Calculate the target cumulative journey based on the cumulative journey and path distance, retrieve the upper bound sampling index, and calculate the distance-location ratio, propagation arrival time, and propagation time difference;

[0086] Read the left edge distance and right edge distance and verify the corresponding valid measurement closed interval. Record the valid reference section sampling index of the propagation that has passed the verification as the valid reference section sampling index of the edge distance. Calculate the belt lateral offset and the propagation alignment lateral offset.

[0087] The combined resolution is calculated based on the resolution of the left and right distance sensors, and then normalized by the inverse of the combined resolution to form the representative offset contribution share across the measurement positions.

[0088] The lateral offset is aligned by weighted cumulative propagation based on the offset contribution share across measurement locations to form the representative offset across measurement locations.

[0089] Calculate the cross-position offset span, cross-position cumulative offset change, and net lateral change based on the propagation alignment lateral offset;

[0090] The propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change are summarized to generate a belt offset measurement record, which is then output in ascending order of edge distance effective reference section sampling index.

[0091] By establishing multiple measurement locations, longitudinal and lateral measurement baselines in the straight conveyor measurement section, and obtaining path distance, effective measurement range, resolution, and baseline edge distance, the problem of lack of unified spatial reference and different sensor installation conditions between different measurement locations is solved. By uniformly sampling the longitudinal conveying speed and left and right edge distances, and establishing sampling indices for cumulative travel, belt entity baseline cross-section, and propagation effective baseline cross-section, the problem of "different measurement locations measuring different belt material at the same time" caused by continuous belt movement is solved. Furthermore, by using target cumulative travel, upper limit sampling index, distance-position ratio, propagation arrival time, and propagation time difference, the measurement results of the same entity cross-section at various locations are propagated and aligned, reducing object misalignment caused by comparing only at the same time or with a fixed delay. By filtering the effective measurement closed interval, data exceeding the effective measurement range of the sensor is reduced from entering subsequent calculations. The contribution share is formed by sensor resolution and weighted for the propagation alignment lateral offset, avoiding mechanically treating locations with different measurement capabilities the same. Finally, the offset span, cumulative offset change, and lateral net change are used to describe the offset evolution of the same entity cross-section on the conveyor path from different perspectives. Compared with common single-point deviation detection or multiple sensors making independent judgments, the beneficial effect of this solution is that it can not only obtain "how much deviation" at a certain position, but also maintain the physical correspondence between multiple positions of the same entity cross section, forming a comprehensive offset result and change result with the meaning of the propagation process, thereby improving the comparability, traceability and engineering interpretability of multi-position measurement results.

[0092] Reference Figure 1 The process of establishing and numbering multiple intelligent sensor measurement positions in the straight conveying measurement section, establishing longitudinal and transverse measurement baselines for the conveying path, acquiring the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and baseline edge distances, forming effective measurement closed intervals, and calculating the baseline distance difference, specifically includes:

[0093] In the linear conveyor measurement section, at least two intelligent sensor measurement positions are established sequentially along the belt running direction. The measurement positions are numbered according to their arrangement order. The total number of intelligent sensor measurement positions included in the current measurement is counted, and a longitudinal baseline of the conveyor path is established that runs through all measurement positions.

[0094] Set the first measurement position as the longitudinal distance reference, read the path distance of each measurement position relative to the first measurement position in sequence along the longitudinal reference line of the conveying path, record the path distance of the first measurement position as zero, and record the path distance of subsequent measurement positions in ascending order of measurement position number.

[0095] Facing the direction of belt running, the right side of the belt is defined as the positive direction and the left side as the negative direction. At each measurement position, a transverse measurement baseline perpendicular to the longitudinal baseline of the conveying path is established. On each transverse measurement baseline, a left fixed measurement baseline and a right fixed measurement baseline are established relative to the conveying device. The distance from the left fixed measurement baseline to the left edge of the belt and the distance from the right edge of the belt to the right fixed measurement baseline are read along the corresponding transverse measurement baseline.

[0096] Read the effective lower limit and effective upper limit of the left and right distance sensors at each measurement position respectively. The effective lower limit and effective upper limit of the same distance sensor constitute the effective measurement closed interval of the corresponding distance sensor. Compare the effective lower limit and effective upper limit of the same distance sensor. The effective measurement boundary condition is that the effective lower limit is a non-negative value and the effective lower limit is less than the corresponding effective upper limit.

[0097] When the left or right distance sensor at any measurement position does not meet the valid measurement boundary conditions, the current measurement is determined to be invalid and the generation of the belt offset measurement record corresponding to the current measurement is stopped. When the left and right distance sensors at all measurement positions meet the valid measurement boundary conditions, under the centered calibration state where the longitudinal center line of the belt coincides with the longitudinal reference line of the conveying path, the left reference edge distance and right reference edge distance of each measurement position are read and compared with the effective lower limit and effective upper limit of the corresponding distance sensor respectively.

[0098] When any left or right reference edge distance is outside the effective measurement closed interval of the corresponding distance sensor, the current measurement is deemed invalid and the generation of belt offset measurement records is stopped. When all left and right reference edge distances are within the effective measurement closed interval of their respective corresponding distance sensors, the left reference edge distance is subtracted from the right reference edge distance for each measurement position to form the reference distance difference of the current measurement position.

[0099] Read the resolution of the left and right distance sensors at each measurement position respectively, compare the resolution of the left and right distance sensors at each measurement position with zero respectively, and check whether the resolution of the left and right distance sensors at each measurement position is positive.

[0100] When the resolution of the left and right distance sensors at all measurement positions is positive, the resolution of the left and right distance sensors at each measurement position is written into the cross-measurement position offset contribution share calculation data. When the resolution of the left or right distance sensor at any measurement position is not positive, the current measurement is determined to be invalid and the generation of belt offset measurement records is stopped.

[0101] Reference Figure 2 The process of uniformly sampling the longitudinal conveyor speed measurement, left edge distance, and right edge distance; calculating the constant equivalent longitudinal conveyor speed and cumulative travel; establishing the belt physical reference section, reference section sampling index, and reference section sampling time; and filtering the propagation of valid reference section sampling indexes specifically includes:

[0102] Read the unified sampling period, the start sampling time of the current sampling batch, and the sampling quantity of the current sampling batch. Check whether the unified sampling period is a positive real number and check whether the sampling quantity of the current sampling batch is not less than two. Use the fact that the unified sampling period is a positive real number and the sampling quantity of the current sampling batch is not less than two as the sampling configuration conditions.

[0103] When the current sampling batch does not meet the sampling configuration conditions, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When the current sampling batch meets the sampling configuration conditions, the unified sampling time is generated sequentially according to the unified sampling period starting from the initial sampling time of the current sampling batch. The discrete sampling index is established from zero according to the time sequence. The unified longitudinal conveying speed measurement value of the physical material on the belt surface, as well as the left edge distance and right edge distance of all measurement positions are uniformly written into the corresponding unified sampling time.

[0104] For each uniform sampling moment, a straight conveying measurement section with the same longitudinal conveying speed is used at each cross-sectional position of the solid material on the belt surface. The uniform longitudinal conveying speed measurement value obtained directly from the solid material on the belt surface is read. The direction of the uniform longitudinal conveying speed measurement value is specified according to the increasing direction of the measurement position number. Each uniform longitudinal conveying speed measurement value is checked one by one to see if it is non-negative.

[0105] When there is a negative uniform longitudinal conveying speed measurement value, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When all uniform longitudinal conveying speed measurements are non-negative, for every two adjacent uniform sampling times, the time interval between the two adjacent uniform sampling times is recorded as the corresponding sampling interval. The uniform longitudinal conveying speed measurement value of the previous uniform sampling time is added to the uniform longitudinal conveying speed measurement value of the next uniform sampling time, and then the result is divided by two to calculate the constant equivalent longitudinal conveying speed within the corresponding sampling interval.

[0106] The cumulative travel distance at the start sampling time of the current sampling batch is recorded as zero. For each sampling interval after the start sampling time of the current sampling batch, the corresponding constant equivalent longitudinal transmission speed is multiplied by the unified sampling period to calculate the longitudinal propagation distance of the corresponding sampling interval. Then, the longitudinal propagation distance of the corresponding sampling interval is added to the cumulative travel distance of the previous unified sampling time to calculate the cumulative travel distance corresponding to each unified sampling time, forming a cumulative travel distance sequence arranged according to the unified sampling time. The propagation arrival time within the corresponding sampling interval is processed according to the constant equivalent longitudinal transmission speed of the corresponding sampling interval.

[0107] For each unified sampling time within the current sampling batch, the discrete sampling index corresponding to the unified sampling time is recorded as the reference section sampling index, the cross-sectional position of the belt surface material at the first measurement position corresponding to the unified sampling time is recorded as the belt entity reference section of the corresponding reference section sampling index, the unified sampling time is recorded as the reference section sampling time, and the cumulative travel corresponding to the reference section sampling time is recorded.

[0108] For each reference section sampling index corresponding to the belt entity reference section before the last unified sampling time, compare the cumulative travel of the previous unified sampling time with the cumulative travel of the current reference section sampling time. Only retain the reference section sampling index corresponding to the cumulative travel of the previous unified sampling time being greater than the cumulative travel of the current reference section sampling time. Then, subtract the cumulative travel of the corresponding reference section sampling time from the cumulative travel of the last unified sampling time of the current sampling batch to calculate the remaining cumulative travel. Compare the remaining cumulative travel with the path distance of the last measurement position relative to the first measurement position. Only retain the reference section sampling index with the remaining cumulative travel not less than the path distance of the last measurement position relative to the first measurement position. Record the retained reference section sampling index as the propagation effective reference section sampling index, and form the propagation effective reference section sampling index set from the propagation effective reference section sampling index.

[0109] When the set of effective reference section sampling indexes for propagation is empty, the generation of belt offset measurement records for the current sampling batch is stopped. When the set of effective reference section sampling indexes for propagation is not empty, the propagation arrival time, propagation time difference, and propagation alignment lateral offset of each measurement position are calculated one by one according to the effective reference section sampling indexes for propagation.

[0110] Reference Figure 3 The step of calculating the target cumulative distance based on the cumulative distance traveled and path distance, retrieving the upper bound sampling index, and calculating the distance-position ratio, propagation arrival time, and propagation time difference specifically includes:

[0111] For each belt entity reference section corresponding to the sampling index of the effective reference section, the cumulative travel time of the reference section corresponding to the current belt entity reference section is read. For each measurement position from the second measurement position to the last measurement position, the cumulative travel time of the reference section corresponding to the current belt entity reference section at the sampling time is added to the path distance of the current measurement position relative to the first measurement position to calculate the target cumulative travel time when the current belt entity reference section propagates to the current measurement position.

[0112] Read the sampling time of the reference section corresponding to the current reference section of the belt entity, record the sampling time of the reference section as the propagation arrival time of the first measurement position, and record the propagation time difference of the current reference section of the belt entity at the first measurement position as zero;

[0113] For each measurement position from the second measurement position to the last measurement position, starting from the cumulative travel sequence after the sampling time of the corresponding reference section, each cumulative travel is compared with the corresponding target cumulative travel one by one according to the sampling time sequence, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is retrieved, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is recorded as the upper bound sampling index.

[0114] Read the cumulative travel distance corresponding to the upper bound sampling index and the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the remaining propagation distance of the target cumulative travel distance relative to the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the cumulative travel distance increment between the target and the previous discrete sampling index. Divide the remaining propagation distance of the target cumulative travel distance relative to the previous discrete sampling index by the cumulative travel distance increment between the target and the previous discrete sampling index to calculate the distance-position ratio.

[0115] Read the unified sampling time corresponding to the upper bound sampling index and the unified sampling time corresponding to the previous discrete sampling index. Subtract the unified sampling time corresponding to the previous discrete sampling index from the unified sampling time corresponding to the upper bound sampling index to calculate the sampling time length between the upper bound sampling index and the previous discrete sampling index. Multiply the sampling time length between the upper bound sampling index and the previous discrete sampling index by the distance-position ratio to calculate the time increment. Add the time increment to the unified sampling time corresponding to the previous discrete sampling index to calculate the propagation arrival time of the belt entity reference section to the corresponding measurement position.

[0116] For each measurement position from the second measurement position to the last measurement position, the propagation time difference from the first measurement position to the corresponding measurement position is calculated by subtracting the sampling time of the reference section corresponding to the current belt entity reference section from the corresponding propagation arrival time.

[0117] Reference Figure 3 The process of reading the left edge distance and right edge distance and verifying the corresponding valid measurement closed interval, recording the valid propagation reference section sampling index that has passed the verification as the edge distance valid reference section sampling index, and calculating the belt lateral offset and propagation alignment lateral offset specifically includes:

[0118] For each belt entity reference section corresponding to the effective reference section sampling index, read the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section, and record the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section as the left edge distance and right edge distance of the first measurement position actually participating in the lateral offset reconstruction.

[0119] For each measurement position from the second measurement position to the last measurement position, the corresponding distance-position ratio is read. When the distance-position ratio is equal to one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index are read. When the distance-position ratio is greater than zero and less than one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index and the left edge distance and right edge distance at the unified sampling time corresponding to the previous discrete sampling index are read respectively. The left edge distance and right edge distance read according to the distance-position ratio are recorded as the left edge distance and right edge distance of the corresponding measurement position that actually participate in the lateral offset reconstruction.

[0120] For each left edge distance that actually participates in the lateral offset reconstruction, the current left edge distance is compared with the effective lower limit and effective upper limit of the corresponding left distance sensor. For each right edge distance that actually participates in the lateral offset reconstruction, the current right edge distance is compared with the effective lower limit and effective upper limit of the corresponding right distance sensor.

[0121] When all measured positions are within the effective measurement closed interval of their respective distance sensors, the left and right edge distances that actually participate in the lateral offset reconstruction are recorded as the edge distance effective reference section sampling index. When any left or right edge distance that actually participates in the lateral offset reconstruction is outside the effective measurement closed interval of the corresponding distance sensor, the measurement result of the current propagation effective reference section sampling index is determined to be invalid edge distance. The calculation of belt lateral offset, propagation alignment lateral offset, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, and lateral net change corresponding to the current propagation effective reference section sampling index is stopped, and the generation of the corresponding belt offset measurement record is stopped.

[0122] For each group of actual left and right edge distances participating in the lateral offset reconstruction corresponding to the effective reference section sampling index of edge distance, the left edge distance is successively subtracted from the right edge distance and the reference distance difference formed by the corresponding measurement position in the centering calibration state. The difference is then divided by two to calculate the belt lateral offset at the corresponding measurement position at the corresponding unified sampling time. Belt lateral offsets greater than zero are recorded as belt lateral offset to the right, belt lateral offsets less than zero are recorded as belt lateral offset to the left, and belt lateral offsets equal to zero are recorded as no lateral offset formed relative to the centering calibration state at the corresponding measurement position and unified sampling time.

[0123] The left edge distance and right edge distance corresponding to the sampling time of the reference section are read at the first measurement position. The belt lateral offset corresponding to the sampling time of the reference section is calculated by subtracting the right edge distance and the difference between the reference distance at the first measurement position from the left edge distance and then dividing by two.

[0124] For each measurement position from the second to the last measurement position, when the distance-position ratio is equal to one, the belt lateral offset at the unified sampling time corresponding to the upper bound sampling index is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index. When the distance-position ratio is greater than zero and less than one, the belt lateral offset corresponding to each of the two adjacent unified sampling times is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the previous discrete sampling index of the upper bound sampling index, as well as the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index.

[0125] For the first measurement position, the lateral offset of the belt corresponding to the sampling time of the reference section is directly recorded as the propagation alignment lateral offset of the current belt entity reference section at the first measurement position. For the second to the last measurement position, when the distance-position ratio is equal to one, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is directly recorded as the propagation alignment lateral offset of the corresponding measurement position. When the distance-position ratio is greater than zero and less than one, the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index is calculated by subtracting the distance-position ratio from one. The lateral offset of the belt at the unified sampling time corresponding to the previous discrete sampling index is multiplied by the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index. Then, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is multiplied by the distance-position ratio. The two products are added together to calculate the propagation alignment lateral offset of the current belt entity reference section at the corresponding measurement position.

[0126] Reference Figure 4 The step of calculating the combined resolution based on the resolutions of the left and right distance sensors, and normalizing it by the inverse of the combined resolution to form the representative offset contribution share across measurement positions, specifically includes:

[0127] For all measurement locations, the resolutions of the left and right distance sensors, which serve as the data for calculating the contribution share of offset across measurement locations, are read respectively.

[0128] For each measurement location, the resolution of the left distance sensor and the resolution of the right distance sensor at the current measurement location are added together to calculate the combined resolution of the left and right distance sensors at the current measurement location.

[0129] For each measurement location, the reciprocal of the combined resolution of the left and right distance sensors at the current measurement location is taken to calculate the resolution weight base value corresponding to the current measurement location;

[0130] The resolution weight base values ​​corresponding to all measurement locations are added together sequentially to calculate the total resolution weight base values ​​of all measurement locations;

[0131] For each measurement location, the resolution weight base value corresponding to the current measurement location is divided by the sum of the resolution weight base values ​​of all measurement locations to calculate the cross-measurement location representative offset contribution share of the current measurement location.

[0132] Reference Figure 4 The step of weighting and accumulating the lateral offset according to the contribution share of the offset across the measurement location to form the representative offset across the measurement location specifically includes:

[0133] For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, read the propagation alignment lateral offset of all measurement positions and the cross-measurement position representative offset contribution share of each measurement position in ascending order of measurement position number;

[0134] For each measurement location, the propagation alignment lateral offset of the current measurement location is multiplied by the cross-measurement location representative offset contribution share corresponding to the current measurement location to calculate the representative offset component corresponding to the current measurement location.

[0135] The representative offset components corresponding to all measurement positions are added together in sequence to calculate the representative offset across the measurement positions of the current belt entity reference section. The representative offset across the measurement positions that is greater than zero is recorded as the result of being biased to the right side of the belt, and the representative offset across the measurement positions that is less than zero is recorded as the result of being biased to the left side of the belt.

[0136] Reference Figure 4 The calculation of the cross-position offset span, the cumulative cross-position offset change, and the net lateral change based on the propagation alignment lateral offset specifically includes:

[0137] For each belt entity reference section corresponding to the effective reference section sampling index at each edge distance, read the propagation alignment lateral offset of all measurement positions, filter out the maximum and minimum propagation alignment lateral offsets from all measurement positions, subtract the minimum propagation alignment lateral offset from the maximum propagation alignment lateral offset, and calculate the cross-position offset span of the current belt entity reference section.

[0138] According to the ascending order of the measurement position numbers, for each group of adjacent measurement positions, the propagation alignment lateral offset of the next measurement position is subtracted from the propagation alignment lateral offset of the previous measurement position. The absolute value of the difference is taken, and then the absolute values ​​corresponding to all adjacent measurement positions are added together to calculate the cumulative offset change of the current belt entity reference section across the position.

[0139] Subtract the lateral offset of the propagation alignment from the first measurement position from the lateral offset of the propagation alignment at the last measurement position to calculate the net lateral change of the current belt entity reference section from the first measurement position to the last measurement position. Record the net lateral change that is greater than zero as the net change of the current belt entity reference section to the right lateral side of the belt after propagation from the first measurement position to the last measurement position, and record the net lateral change that is less than zero as the net change of the current belt entity reference section to the left lateral side of the belt after propagation from the first measurement position to the last measurement position.

[0140] Reference Figure 4 The process of summarizing the propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change to generate a belt offset measurement record, and outputting it in ascending order of edge distance effective reference section sampling index, specifically includes:

[0141] For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, the propagation alignment lateral offset of each measurement position is read in ascending order of measurement position number, and all propagation alignment lateral offsets are arranged in ascending order of measurement position number to generate multi-position propagation alignment offset ordered data corresponding to the current belt entity reference section.

[0142] According to the ascending order of the measurement position numbers, the propagation time difference of the current belt physical reference section from the first measurement position to each measurement position is read sequentially, and all propagation time differences are arranged in ascending order of the measurement position numbers to generate ordered data of the propagation time difference corresponding to the current belt physical reference section.

[0143] The system sequentially reads the current edge distance effective reference section sampling index, multi-position propagation alignment offset ordered data, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, lateral net change, and propagation time difference ordered data. It then combines these data according to the correspondence of the same belt entity reference section to generate the belt offset measurement record corresponding to the current edge distance effective reference section sampling index.

[0144] The belt offset measurement records are arranged and output in ascending order of the effective reference section sampling index based on the edge distance.

[0145] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0146] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for measuring belt offset based on intelligent sensors, characterized in that, include: In the straight conveying measurement section, establish and number multiple intelligent sensor measurement positions, establish longitudinal and transverse measurement baselines for the conveying path, obtain the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and baseline edge distances, form effective measurement closed intervals, and calculate the baseline distance difference; Uniform sampling is performed on the longitudinal conveying speed measurement, left edge distance, and right edge distance. The constant equivalent longitudinal conveying speed and cumulative travel are calculated. The belt physical reference section, reference section sampling index, and reference section sampling time are established. The effective reference section sampling index is then selected for propagation. Calculate the target cumulative journey based on the cumulative journey and path distance, retrieve the upper bound sampling index, and calculate the distance-location ratio, propagation arrival time, and propagation time difference; Read the left edge distance and right edge distance and verify the corresponding valid measurement closed interval. Record the valid reference section sampling index of the propagation that has passed the verification as the valid reference section sampling index of the edge distance. Calculate the belt lateral offset and the propagation alignment lateral offset. The combined resolution is calculated based on the resolution of the left and right distance sensors, and then normalized by the inverse of the combined resolution to form the representative offset contribution share across the measurement positions. The lateral offset is aligned by weighted cumulative propagation based on the offset contribution share across measurement locations to form the representative offset across measurement locations. Calculate the cross-position offset span, cross-position cumulative offset change, and net lateral change based on the propagation alignment lateral offset; The propagation alignment lateral offset, propagation time difference, representative offset across measurement positions, offset span across positions, cumulative offset change across positions, and net lateral change are summarized to generate a belt offset measurement record, which is then output in ascending order of edge distance effective reference section sampling index.

2. The belt offset measurement method based on a smart sensor according to claim 1, characterized in that, The process of establishing and numbering multiple intelligent sensor measurement positions in the straight conveying measurement section, establishing longitudinal and transverse measurement baselines for the conveying path, acquiring the effective upper and lower limits and resolutions of the path distance, left and right distance sensors, and baseline edge distances, forming an effective measurement closed interval, and calculating the baseline distance difference specifically includes: In the linear conveyor measurement section, at least two intelligent sensor measurement positions are established sequentially along the belt running direction. The measurement positions are numbered according to their arrangement order. The total number of intelligent sensor measurement positions included in the current measurement is counted, and a longitudinal baseline of the conveyor path is established that runs through all measurement positions. Set the first measurement position as the longitudinal distance reference, read the path distance of each measurement position relative to the first measurement position in sequence along the longitudinal reference line of the conveying path, record the path distance of the first measurement position as zero, and record the path distance of subsequent measurement positions in ascending order of measurement position number. Facing the direction of belt running, the right side of the belt is defined as the positive direction and the left side as the negative direction. At each measurement position, a transverse measurement baseline perpendicular to the longitudinal baseline of the conveying path is established. On each transverse measurement baseline, a left fixed measurement baseline and a right fixed measurement baseline are established relative to the conveying device. The distance from the left fixed measurement baseline to the left edge of the belt and the distance from the right edge of the belt to the right fixed measurement baseline are read along the corresponding transverse measurement baseline. Read the effective lower limit and effective upper limit of the left and right distance sensors at each measurement position respectively. The effective lower limit and effective upper limit of the same distance sensor constitute the effective measurement closed interval of the corresponding distance sensor. Compare the effective lower limit and effective upper limit of the same distance sensor. The effective measurement boundary condition is that the effective lower limit is a non-negative value and the effective lower limit is less than the corresponding effective upper limit. When the left or right distance sensor at any measurement position does not meet the valid measurement boundary conditions, the current measurement is determined to be invalid and the generation of the belt offset measurement record corresponding to the current measurement is stopped. When the left and right distance sensors at all measurement positions meet the valid measurement boundary conditions, under the centered calibration state where the longitudinal center line of the belt coincides with the longitudinal reference line of the conveying path, the left reference edge distance and right reference edge distance of each measurement position are read and compared with the effective lower limit and effective upper limit of the corresponding distance sensor respectively. When any left or right reference edge distance is outside the effective measurement closed interval of the corresponding distance sensor, the current measurement is deemed invalid and the generation of belt offset measurement records is stopped. When all left and right reference edge distances are within the effective measurement closed interval of their respective corresponding distance sensors, the left reference edge distance is subtracted from the right reference edge distance for each measurement position to form the reference distance difference of the current measurement position. Read the resolution of the left and right distance sensors at each measurement position respectively, compare the resolution of the left and right distance sensors at each measurement position with zero respectively, and check whether the resolution of the left and right distance sensors at each measurement position is positive. When the resolution of the left and right distance sensors at all measurement positions is positive, the resolution of the left and right distance sensors at each measurement position is written into the cross-measurement position offset contribution share calculation data. When the resolution of the left or right distance sensor at any measurement position is not positive, the current measurement is determined to be invalid and the generation of belt offset measurement records is stopped.

3. The belt offset measurement method based on a smart sensor according to claim 2, characterized in that, The process of uniformly sampling the longitudinal conveyor speed measurement, left edge distance, and right edge distance; calculating the constant equivalent longitudinal conveyor speed and cumulative travel; establishing the belt physical reference section, reference section sampling index, and reference section sampling time; and filtering the effective reference section sampling index for propagation specifically includes: Read the unified sampling period, the start sampling time of the current sampling batch, and the sampling quantity of the current sampling batch. Check whether the unified sampling period is a positive real number and check whether the sampling quantity of the current sampling batch is not less than two. Use the fact that the unified sampling period is a positive real number and the sampling quantity of the current sampling batch is not less than two as the sampling configuration conditions. When the current sampling batch does not meet the sampling configuration conditions, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When the current sampling batch meets the sampling configuration conditions, the unified sampling time is generated sequentially according to the unified sampling period starting from the initial sampling time of the current sampling batch. The discrete sampling index is established from zero according to the time sequence. The unified longitudinal conveying speed measurement value of the physical material on the belt surface, as well as the left edge distance and right edge distance of all measurement positions are uniformly written into the corresponding unified sampling time. For each uniform sampling moment, a straight conveying measurement section with the same longitudinal conveying speed is used at each cross-sectional position of the solid material on the belt surface. The uniform longitudinal conveying speed measurement value obtained directly from the solid material on the belt surface is read. The direction of the uniform longitudinal conveying speed measurement value is specified according to the increasing direction of the measurement position number. Each uniform longitudinal conveying speed measurement value is checked one by one to see if it is non-negative. When there is a negative uniform longitudinal conveying speed measurement value, the current sampling batch is determined to be invalid and the generation of belt offset measurement records is stopped. When all uniform longitudinal conveying speed measurements are non-negative, for every two adjacent uniform sampling times, the time interval between the two adjacent uniform sampling times is recorded as the corresponding sampling interval. The uniform longitudinal conveying speed measurement value of the previous uniform sampling time is added to the uniform longitudinal conveying speed measurement value of the next uniform sampling time, and then the result is divided by two to calculate the constant equivalent longitudinal conveying speed within the corresponding sampling interval. The cumulative travel distance at the start sampling time of the current sampling batch is recorded as zero. For each sampling interval after the start sampling time of the current sampling batch, the corresponding constant equivalent longitudinal transmission speed is multiplied by the unified sampling period to calculate the longitudinal propagation distance of the corresponding sampling interval. Then, the longitudinal propagation distance of the corresponding sampling interval is added to the cumulative travel distance of the previous unified sampling time to calculate the cumulative travel distance corresponding to each unified sampling time, forming a cumulative travel distance sequence arranged according to the unified sampling time. The propagation arrival time within the corresponding sampling interval is processed according to the constant equivalent longitudinal transmission speed of the corresponding sampling interval. For each unified sampling time within the current sampling batch, the discrete sampling index corresponding to the unified sampling time is recorded as the reference section sampling index, the cross-sectional position of the belt surface material at the first measurement position corresponding to the unified sampling time is recorded as the belt entity reference section of the corresponding reference section sampling index, the unified sampling time is recorded as the reference section sampling time, and the cumulative travel corresponding to the reference section sampling time is recorded. For each reference section sampling index corresponding to the belt entity reference section before the last unified sampling time, compare the cumulative travel of the previous unified sampling time with the cumulative travel of the current reference section sampling time. Only retain the reference section sampling index corresponding to the cumulative travel of the previous unified sampling time being greater than the cumulative travel of the current reference section sampling time. Then, subtract the cumulative travel of the corresponding reference section sampling time from the cumulative travel of the last unified sampling time of the current sampling batch to calculate the remaining cumulative travel. Compare the remaining cumulative travel with the path distance of the last measurement position relative to the first measurement position. Only retain the reference section sampling index with the remaining cumulative travel not less than the path distance of the last measurement position relative to the first measurement position. Record the retained reference section sampling index as the propagation effective reference section sampling index, and form the propagation effective reference section sampling index set from the propagation effective reference section sampling index. When the set of effective reference section sampling indexes for propagation is empty, the generation of belt offset measurement records for the current sampling batch is stopped. When the set of effective reference section sampling indexes for propagation is not empty, the propagation arrival time, propagation time difference, and propagation alignment lateral offset of each measurement position are calculated one by one according to the effective reference section sampling indexes for propagation.

4. The belt offset measurement method based on a smart sensor according to claim 3, characterized in that, The process of calculating the target cumulative journey based on the cumulative journey and path distance, retrieving the upper bound sampling index, and calculating the distance-location ratio, propagation arrival time, and propagation time difference specifically includes: For each belt entity reference section corresponding to the sampling index of the effective reference section, the cumulative travel time of the reference section corresponding to the current belt entity reference section is read. For each measurement position from the second measurement position to the last measurement position, the cumulative travel time of the reference section corresponding to the current belt entity reference section at the sampling time is added to the path distance of the current measurement position relative to the first measurement position to calculate the target cumulative travel time when the current belt entity reference section propagates to the current measurement position. Read the sampling time of the reference section corresponding to the current reference section of the belt entity, record the sampling time of the reference section as the propagation arrival time of the first measurement position, and record the propagation time difference of the current reference section of the belt entity at the first measurement position as zero; For each measurement position from the second measurement position to the last measurement position, starting from the cumulative travel sequence after the sampling time of the corresponding reference section, each cumulative travel is compared with the corresponding target cumulative travel one by one according to the sampling time sequence, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is retrieved, and the discrete sampling index that first reaches or exceeds the corresponding target cumulative travel is recorded as the upper bound sampling index. Read the cumulative travel distance corresponding to the upper bound sampling index and the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the remaining propagation distance of the target cumulative travel distance relative to the cumulative travel distance corresponding to the previous discrete sampling index. Subtract the cumulative travel distance corresponding to the previous discrete sampling index from the target cumulative travel distance to calculate the cumulative travel distance increment between the target and the previous discrete sampling index. Divide the remaining propagation distance of the target cumulative travel distance relative to the previous discrete sampling index by the cumulative travel distance increment between the target and the previous discrete sampling index to calculate the distance-position ratio. Read the unified sampling time corresponding to the upper bound sampling index and the unified sampling time corresponding to the previous discrete sampling index. Subtract the unified sampling time corresponding to the previous discrete sampling index from the unified sampling time corresponding to the upper bound sampling index to calculate the sampling time length between the upper bound sampling index and the previous discrete sampling index. Multiply the sampling time length between the upper bound sampling index and the previous discrete sampling index by the distance-position ratio to calculate the time increment. Add the time increment to the unified sampling time corresponding to the previous discrete sampling index to calculate the propagation arrival time of the belt entity reference section to the corresponding measurement position. For each measurement position from the second measurement position to the last measurement position, the propagation time difference from the first measurement position to the corresponding measurement position is calculated by subtracting the sampling time of the reference section corresponding to the current belt entity reference section from the corresponding propagation arrival time.

5. The belt offset measurement method based on a smart sensor according to claim 4, characterized in that, The process of reading the left and right edge distances and verifying the corresponding valid measurement closed intervals, recording the verified valid propagation reference section sampling index as the edge distance valid reference section sampling index, and calculating the belt lateral offset and propagation alignment lateral offset specifically includes: For each belt entity reference section corresponding to the effective reference section sampling index, read the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section, and record the left edge distance and right edge distance obtained by the first measurement position at the sampling time of the corresponding reference section as the left edge distance and right edge distance of the first measurement position actually participating in the lateral offset reconstruction. For each measurement position from the second measurement position to the last measurement position, the corresponding distance-position ratio is read. When the distance-position ratio is equal to one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index are read. When the distance-position ratio is greater than zero and less than one, the left edge distance and right edge distance at the unified sampling time corresponding to the upper bound sampling index and the left edge distance and right edge distance at the unified sampling time corresponding to the previous discrete sampling index are read respectively. The left edge distance and right edge distance read according to the distance-position ratio are recorded as the left edge distance and right edge distance of the corresponding measurement position that actually participate in the lateral offset reconstruction. For each left edge distance that actually participates in the lateral offset reconstruction, the current left edge distance is compared with the effective lower limit and effective upper limit of the corresponding left distance sensor. For each right edge distance that actually participates in the lateral offset reconstruction, the current right edge distance is compared with the effective lower limit and effective upper limit of the corresponding right distance sensor. When all measured positions are within the effective measurement closed interval of their respective distance sensors, the left and right edge distances that actually participate in the lateral offset reconstruction are recorded as the edge distance effective reference section sampling index. When any left or right edge distance that actually participates in the lateral offset reconstruction is outside the effective measurement closed interval of the corresponding distance sensor, the measurement result of the current propagation effective reference section sampling index is determined to be invalid edge distance. The calculation of belt lateral offset, propagation alignment lateral offset, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, and lateral net change corresponding to the current propagation effective reference section sampling index is stopped, and the generation of the corresponding belt offset measurement record is stopped. For each group of actual left and right edge distances participating in the lateral offset reconstruction corresponding to the effective reference section sampling index of edge distance, the left edge distance is successively subtracted from the right edge distance and the reference distance difference formed by the corresponding measurement position in the centering calibration state. The difference is then divided by two to calculate the belt lateral offset at the corresponding measurement position at the corresponding unified sampling time. Belt lateral offsets greater than zero are recorded as belt lateral offset to the right, belt lateral offsets less than zero are recorded as belt lateral offset to the left, and belt lateral offsets equal to zero are recorded as no lateral offset formed relative to the centering calibration state at the corresponding measurement position and unified sampling time. The left edge distance and right edge distance corresponding to the sampling time of the reference section are read at the first measurement position. The belt lateral offset corresponding to the sampling time of the reference section is calculated by subtracting the right edge distance and the difference between the reference distance at the first measurement position from the left edge distance and then dividing by two. For each measurement position from the second to the last measurement position, when the distance-position ratio is equal to one, the belt lateral offset at the unified sampling time corresponding to the upper bound sampling index is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index. When the distance-position ratio is greater than zero and less than one, the belt lateral offset corresponding to each of the two adjacent unified sampling times is calculated according to the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the previous discrete sampling index of the upper bound sampling index, as well as the difference between the left edge distance, the right edge distance, and the reference distance of the corresponding measurement position at the unified sampling time corresponding to the upper bound sampling index. For the first measurement position, the lateral offset of the belt corresponding to the sampling time of the reference section is directly recorded as the propagation alignment lateral offset of the current belt entity reference section at the first measurement position. For the second to the last measurement position, when the distance-position ratio is equal to one, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is directly recorded as the propagation alignment lateral offset of the corresponding measurement position. When the distance-position ratio is greater than zero and less than one, the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index is calculated by subtracting the distance-position ratio from one. The lateral offset of the belt at the unified sampling time corresponding to the previous discrete sampling index is multiplied by the ratio coefficient corresponding to the unified sampling time of the previous discrete sampling index. Then, the lateral offset of the belt at the unified sampling time corresponding to the upper bound sampling index is multiplied by the distance-position ratio. The two products are added together to calculate the propagation alignment lateral offset of the current belt entity reference section at the corresponding measurement position.

6. The belt offset measurement method based on a smart sensor according to claim 5, characterized in that, The process of calculating the combined resolution based on the resolutions of the left and right distance sensors, and then normalizing it by the inverse of the combined resolution to form the representative offset contribution share across measurement positions, specifically includes: For all measurement locations, the resolutions of the left and right distance sensors, which serve as the data for calculating the contribution share of offset across measurement locations, are read respectively. For each measurement location, the resolution of the left distance sensor and the resolution of the right distance sensor at the current measurement location are added together to calculate the combined resolution of the left and right distance sensors at the current measurement location. For each measurement location, the reciprocal of the combined resolution of the left and right distance sensors at the current measurement location is taken to calculate the resolution weight base value corresponding to the current measurement location; The resolution weight base values ​​corresponding to all measurement locations are added together sequentially to calculate the total resolution weight base values ​​of all measurement locations; For each measurement location, the resolution weight base value corresponding to the current measurement location is divided by the sum of the resolution weight base values ​​of all measurement locations to calculate the cross-measurement location representative offset contribution share of the current measurement location.

7. The belt offset measurement method based on a smart sensor according to claim 6, characterized in that, The method of weighted cumulative propagation and alignment of lateral offsets based on the offset contribution share across measurement locations to form a representative offset across measurement locations specifically includes: For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, read the propagation alignment lateral offset of all measurement positions and the cross-measurement position representative offset contribution share of each measurement position in ascending order of measurement position number; For each measurement location, the propagation alignment lateral offset of the current measurement location is multiplied by the cross-measurement location representative offset contribution share corresponding to the current measurement location to calculate the representative offset component corresponding to the current measurement location. The representative offset components corresponding to all measurement positions are added together in sequence to calculate the representative offset across the measurement positions of the current belt entity reference section. The representative offset across the measurement positions that is greater than zero is recorded as the result of being biased to the right side of the belt, and the representative offset across the measurement positions that is less than zero is recorded as the result of being biased to the left side of the belt.

8. The belt offset measurement method based on a smart sensor according to claim 7, characterized in that, The calculation of the cross-position offset span, the cumulative cross-position offset change, and the net lateral change based on the propagation alignment lateral offset specifically includes: For each belt entity reference section corresponding to the effective reference section sampling index at each edge distance, read the propagation alignment lateral offset of all measurement positions, filter out the maximum and minimum propagation alignment lateral offsets from all measurement positions, subtract the minimum propagation alignment lateral offset from the maximum propagation alignment lateral offset, and calculate the cross-position offset span of the current belt entity reference section. According to the ascending order of the measurement position numbers, for each group of adjacent measurement positions, the propagation alignment lateral offset of the next measurement position is subtracted from the propagation alignment lateral offset of the previous measurement position. The absolute value of the difference is taken, and then the absolute values ​​corresponding to all adjacent measurement positions are added together to calculate the cumulative offset change of the current belt entity reference section across the position. Subtract the lateral offset of the propagation alignment from the first measurement position from the lateral offset of the propagation alignment at the last measurement position to calculate the net lateral change of the current belt entity reference section from the first measurement position to the last measurement position. Record the net lateral change that is greater than zero as the net change of the current belt entity reference section to the right lateral side of the belt after propagation from the first measurement position to the last measurement position, and record the net lateral change that is less than zero as the net change of the current belt entity reference section to the left lateral side of the belt after propagation from the first measurement position to the last measurement position.

9. A method for measuring belt offset based on a smart sensor according to claim 8, characterized in that, The process of summarizing the lateral offset of the propagation alignment, the propagation time difference, the representative offset across measurement positions, the offset span across positions, the cumulative offset change across positions, and the net lateral change to generate a belt offset measurement record, and outputting it in ascending order of the sampling index of the effective reference section at the edge distance, specifically includes: For each belt entity reference section corresponding to the effective reference section sampling index of the edge distance, the propagation alignment lateral offset of each measurement position is read in ascending order of measurement position number, and all propagation alignment lateral offsets are arranged in ascending order of measurement position number to generate multi-position propagation alignment offset ordered data corresponding to the current belt entity reference section. According to the ascending order of the measurement position numbers, the propagation time difference of the current belt physical reference section from the first measurement position to each measurement position is read sequentially, and all propagation time differences are arranged in ascending order of the measurement position numbers to generate ordered data of the propagation time difference corresponding to the current belt physical reference section. The system sequentially reads the current edge distance effective reference section sampling index, multi-position propagation alignment offset ordered data, cross-measurement position representative offset, cross-position offset span, cross-position cumulative offset change, lateral net change, and propagation time difference ordered data. It then combines these data according to the correspondence of the same belt entity reference section to generate the belt offset measurement record corresponding to the current edge distance effective reference section sampling index. The belt offset measurement records are arranged and output in ascending order of the effective reference section sampling index based on the edge distance.

Images