Method and system for automated operation of warehouse containers based on three-dimensional mechanisms

Abstract

The application relates to the technical field of warehouse logistics management, and discloses a warehouse container automation operation method and system based on a three-dimensional mechanism. The method comprises the following steps: counting operation scene characteristic parameters, calculating a target weight when the characteristic parameters continuously exceed a threshold value, and obtaining a daily weight through five-day linear interpolation smooth transition, scoring candidate storage locations according to the daily weight to select an optimal storage location, and selecting an optimal crane by comprehensively optimizing scheduling according to the moving distance, operation time and queue length. The application improves the self-adaptive capability of the storage location allocation decision to the scene change, the smoothness of the weight adjustment process and the strategy continuity, and the weight configuration convergence and operation efficiency optimization effect based on feedback correction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of warehouse logistics management, and particularly relates to a warehouse container automated operation method and system based on a three-dimensional mechanism. BACKGROUND

[0002] The warehouse container automated operation management method based on a three-dimensional mechanism in the prior art generally includes vehicle arrival identification, storage location allocation, crane scheduling, and container hoisting, etc. The storage location allocation module selects an idle storage location from candidate storage locations according to container attributes and inventory states for allocation. The crane scheduling module issues operation instructions to available cranes according to a task queue. The system obtains basic information such as container types, weights, and sizes by querying a database and performs matching calculation in combination with parameters such as storage location bearing capacities and spatial positions. When a lower container needs to be accessed, the system automatically generates a destacking task to first transfer an upper container to a temporary storage location. After the operation is completed, the system updates container positions and storage location occupancy state records in the inventory database. This kind of method realizes information management of warehouse operation processes to a certain extent.

[0003] However, the prior art has significant deficiencies. The storage location allocation strategy adopts a fixed evaluation weight configuration, that is, the weight coefficients of evaluation factors such as out-of-warehouse frequency, weight distribution, and stacking stability remain constant values during system operation. This static weight configuration cannot adapt to dynamic changes in warehouse operation scenarios. When the warehouse area switches from a low-frequency out-of-warehouse scenario in the off-season to a high-frequency out-of-warehouse scenario in the peak season, the fixed weight still allocates storage locations according to the off-season mode, resulting in that high-frequency out-of-warehouse containers are stored in positions far from the out-of-warehouse passageway, increasing the subsequent carrying distance. When the weight distribution of containers changes from a balanced state to a discrete state, the fixed weight fails to timely improve the priority of the weight balance factor, causing excessive load in local areas and safety hazards. The destacking decision mechanism only considers the immediate transfer path of the upper container and does not comprehensively evaluate the future out-of-warehouse plan of the container and the availability of the temporary storage location. The crane scheduling algorithm performs serial allocation according to the order of tasks without considering the differences in current positions and operation loads of cranes. The system lacks a feedback evaluation and parameter optimization mechanism based on actual operation effects, resulting in that the storage location allocation and scheduling strategy cannot be continuously improved. SUMMARY

[0004] The present application provides a warehouse container automated operation method and system based on a three-dimensional mechanism, which solves the problems in the prior art that the weight configuration of storage location evaluation cannot adapt to dynamic changes in operation scenarios, the weight adjustment process lacks a smooth transition mechanism, resulting in unstable allocation strategies, and there is a lack of weight iterative optimization mechanism based on actual operation effect feedback, improves the self-adaptability of storage location allocation decisions to scenario changes, the smoothness of the weight adjustment process, the strategy continuity, and the convergence of the weight configuration and operation efficiency optimization effect based on feedback correction.

[0005] In a first aspect, the application provides a warehouse container automation operation method based on a three-dimensional mechanism, comprising:

[0006] In step S1, operation task data packets and container information are received, and the warehouse-in and warehouse-out ratio, weight dispersion, and pile overturning pressure index in the past 7 days are counted as scene characteristic parameters.

[0007] In step S2, when any parameter in the scene characteristic parameters exceeds a preset threshold value continuously for 3 days, a target weight vector is calculated, a difference between a current weight vector and the target weight vector is linearly interpolated into a daily adjustment increment for 5 days, and the daily adjustment increment is accumulated to the current weight vector day by day to obtain a daily weight vector.

[0008] In step S3, the warehouse-out probability is predicted according to the container type, the weight deviation, the weight bearing margin, and the distance to the warehouse-out channel of the candidate storage location are queried, the warehouse-out probability, the weight deviation, the weight bearing margin, and the distance are weighted and summed with the daily weight vector to obtain an adaptation score, and the storage location with the highest adaptation score is selected as a target storage location.

[0009] In step S4, the three-dimensional movement distance and operation time of a crane to the container source location and the target storage location are calculated, scheduling optimization is performed according to the movement distance, operation time, and queue length, and the crane with the highest scheduling score is selected to generate a scheduling instruction.

[0010] In a second aspect, the application provides a warehouse container automation operation system based on a three-dimensional mechanism, comprising:

[0011] The receiving module is configured to receive operation task data packets and container information, count the warehouse-in and warehouse-out ratio, weight dispersion, and pile overturning pressure index in the past 7 days as scene characteristic parameters.

[0012] The decomposition module is configured to, when any parameter in the scene characteristic parameters exceeds a preset threshold value continuously for 3 days, calculate a target weight vector, linearly interpolate a difference between a current weight vector and the target weight vector into a daily adjustment increment for 5 days, and accumulate the daily adjustment increment to the current weight vector day by day to obtain a daily weight vector.

[0013] The weighting module is configured to predict the warehouse-out probability according to the container type, query the weight deviation, the weight bearing margin, and the distance to the warehouse-out channel of the candidate storage location, weight and sum the warehouse-out probability, the weight deviation, the weight bearing margin, and the distance with the daily weight vector to obtain an adaptation score, and select the storage location with the highest adaptation score as a target storage location.

[0014] The scheduling module is configured to calculate three-dimensional moving distances and operation times of the crane to a container source position and the target storage location, perform scheduling optimization according to the moving distances, operation times and queue length, and select a crane with the highest scheduling score to generate a scheduling instruction.

[0015] In the technical solution provided in the application, the out-of-warehouse ratio, the weight dispersion degree and the stack pressure index in the past 7 days are counted as scene characteristic parameters, and a scene recognition mechanism based on historical operation data is established. The mechanism can quantitatively reflect the dynamic change trend of the warehouse operation scene. The out-of-warehouse ratio directly describes the out-of-warehouse operation intensity of the warehouse area through the ratio relationship between the total number of out-of-warehouse tasks and the total number of in-warehouse tasks. The weight dispersion degree reveals the dispersion degree of the container weight distribution through the ratio of the standard deviation of the container weight to the average weight. The stack pressure index quantifies the influence of the storage location stack layout on the operation efficiency through the ratio of the number of stack operation times to the total number of out-of-warehouse tasks.

[0016] The three scene characteristic parameters comprehensively represent the operation scene state from different dimensions. When any parameter in the scene characteristic parameters exceeds a preset threshold for 3 consecutive days, a weight adjustment process is triggered. The continuous determination mechanism effectively avoids false triggering caused by single-day data fluctuations. The difference between the current weight vector and the target weight vector is linearly interpolated into a daily adjustment increment for 5 days, and the daily adjustment increments are accumulated to the current weight vector day by day. The smooth transition mechanism uniformly distributes the weight adjustment amount to each day in the 5-day adjustment period, so that the weight configuration of the storage location evaluation model gradually transitions from the current state to the target state instead of jumping, and the same type of containers stored in adjacent time are allocated to storage locations with similar spatial positions due to the continuity of the weight configuration. The weight smoothing adjustment process maintains the stability and predictability of the storage location allocation strategy. The out-of-warehouse probability is predicted according to the container type, and the weight deviation, the remaining load capacity and the distance to the out-of-warehouse channel of the candidate storage location are queried. The evaluation indexes are weighted and summed to obtain an adaptation score. The multi-dimensional evaluation mechanism realizes the adaptive matching of the storage location evaluation strategy to the current operation scene characteristics through the dynamically adjusted weight vector. When the out-of-warehouse ratio is high, the increase of the out-of-warehouse frequency weight makes the containers with high out-of-warehouse probability preferentially allocated to convenient storage locations close to the out-of-warehouse channel. When the weight dispersion degree is large, the increase of the weight balance weight guides the heavy containers to be stored dispersedly to avoid excessive load in local areas. When the stack pressure index is high, the increase of the stack stability weight promotes the containers to be preferentially allocated to storage locations with sufficient remaining load capacity and less overhead shielding. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating labor.

[0018] Figure 1 An embodiment of the warehouse container automation operation method based on a three-dimensional mechanism in the embodiments of the present application is shown in the figure;

[0019] Figure 2 An embodiment of the weight feedback correction algorithm convergence process in the embodiments of the present application is shown in the figure;

[0020] Figure 3 An embodiment of the distribution of each score item of the candidate storage location in the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0021] The embodiments of the present application provide a warehouse container automation operation method and system based on a three-dimensional mechanism. The terms "first", "second", "third", "fourth" and the like (if any) in the description and claims of the present application and the above-mentioned figures are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] For ease of understanding, the specific process of the embodiments of the present application is described below. Please refer to Figure 1 An embodiment of the warehouse container automation operation method based on a three-dimensional mechanism in the embodiments of the present application includes:

[0023] Step S1, receiving operation task data packet and container information, and counting the warehouse in and out ratio, weight dispersion and pile pressure index in the past 7 days as scene characteristic parameters;

[0024] Specifically, the scene feature parameter collection is based on statistical analysis of historical operation data. The warehouse-in / out ratio is obtained by dividing the total number of task records marked as warehouse-out type in the database in the last 7 days by the sum of the total number of warehouse-out and warehouse-in task records. The weight dispersion is obtained by dividing the standard deviation by the arithmetic mean of the weight field of all the containers that have been warehoused. The destacking pressure index reflects the frequency of destacking operation by dividing the number of operation records marked as destacking operation by the total number of warehouse-out tasks. The scene recognition determination mechanism sets three independent threshold detection channels to monitor whether the warehouse-in / out ratio exceeds 0.6 for 3 consecutive days, whether the weight dispersion exceeds 0.3 for 3 consecutive days, and whether the destacking pressure index exceeds 0.2 for 3 consecutive days. If any of the channels is triggered, the scene change identification flag is set.

[0025] Step S2, when any of the scene feature parameters exceeds the preset threshold for 3 consecutive days, calculate the target weight vector, and decompose the difference between the current weight vector and the target weight vector into daily adjustment increments by 5-day linear interpolation. Add the daily adjustment increments to the current weight vector day by day to obtain the daily weight vector.

[0026] Specifically, the weight self-adaptive calculation adopts a gain superposition method. When the warehouse-in / out ratio exceeds 0.6, the difference between the ratio and 0.6 is multiplied by 0.5 to obtain the warehouse-out frequency gain coefficient, which is added to the reference weight 0.35 to form the target warehouse-out frequency weight. The weight dispersion and the destacking pressure index use the same gain calculation logic to generate corresponding target weight components. The weight smoothing transition mechanism subtracts the target weight vector from the current weight vector to obtain four weight differences. Each difference is multiplied by a fixed coefficient 0.2 to obtain a single-day adjustment increment. Through the daily accumulation mechanism, the current weight plus the adjustment increment is multiplied by the number of adjusted days to realize the gradual change of the weight instead of the step change.

[0027] Step S3, predict the warehouse-out probability according to the container type, query the weight deviation, load capacity and distance to the warehouse-out channel of the candidate storage location, and obtain the fitness score by weighted sum of the warehouse-out probability, weight deviation, load capacity and distance and the daily weight vector. Select the storage location with the highest fitness score as the target storage location.

[0028] Specifically, the storage location adaptability evaluation converts the evaluation indexes of the four different dimensions into a unified dimension score item and then performs weighted aggregation. The out-of-warehouse probability is derived from the normalized ratio of the number of out-of-warehouse times of the container type in the last 30 days to the total number of out-of-warehouse times in the last 90 days in the historical database. The weight deviation is quantified by the absolute value of the sum of the weights of the stored containers in the region where the candidate storage location is located minus the average weight of each region in the whole warehouse, and then divided by the average weight, to quantify the unevenness of the weight distribution in the region. The load surplus is directly read from the difference between the design load upper limit field and the current load field in the candidate storage location attribute table. The distance to the out-of-warehouse channel is calculated by the Euclidean norm of the coordinate difference between the candidate storage location and the nearest out-of-warehouse channel. The four original indexes are multiplied by the corresponding weight components in the weight vector of the current day and then summed to form a comprehensive score value. The larger the value, the more suitable the storage location is for the current container storage.

[0029] Step S4, calculate the three-dimensional movement distance and operation time of the crane to the container source location and the target storage location, perform scheduling optimization according to the movement distance, operation time and queue length, and select the crane with the highest scheduling score to generate a scheduling instruction.

[0030] Specifically, the crane scheduling optimization comprehensively considers three dimensions of space cost, time cost and load balance. The space cost is obtained by calculating the Euclidean distance between the current three-dimensional coordinates of the crane and the three-dimensional coordinates of the container source location, and then adding the Euclidean distance between the container source location and the target storage location to obtain the total movement distance. The time cost is obtained by dividing the coordinate difference between the crane and the target storage location in the horizontal, vertical and vertical directions by the corresponding device movement speed parameters, taking the maximum value of the three, and adding the fixed container grabbing and releasing time overhead. The load balance index reflects the device busy degree by reading the number of tasks to be executed in the current operation queue of the crane and dividing by the system set queue capacity upper limit. The three dimension indexes are normalized and weighted summed according to the fixed weight coefficients 0.4, 0.35 and 0.25 to obtain the scheduling score. The crane number with the highest score is selected to be packaged into a scheduling instruction data packet and issued for execution.

[0031] In a specific embodiment, step S1 includes:

[0032] Receiving a job task data packet and parsing the task type field, task region field, job task field and task priority field to obtain the task type, task region number, container target quantity and task priority value;

[0033] Obtaining three-dimensional point cloud data of the container and performing contour extraction processing to obtain three-dimensional size coordinates and attitude inclination angle of the container;

[0034] extracting total number of outbound tasks and total number of inbound tasks in the past 7 days from the historical operation record database, calculating the outbound task total number divided by the sum of the outbound task total number and the inbound task total number to obtain the outbound and inbound ratio;

[0035] extracting weight data of all inbound containers in the past 7 days to calculate the average weight and weight standard deviation, dividing the weight standard deviation by the average weight to obtain the weight dispersion, and calculating the total number of times of unstacking operations in the past 7 days divided by the total number of outbound tasks to obtain the unstacking pressure index.

[0036] Specifically, the operation task data packet is transmitted to the warehouse area management system through a standard interface protocol, and the parsing module locates the field positions according to the pre-defined JSON or XML format structure. The task type field adopts an enumeration value coding mode to store three operation types of inbound, outbound or unstacking. The task area field adopts two-digit coding to map to the warehouse physical partition identification table. The operation task field includes an integer value of the target number of containers and a timestamp range of the completion time window. The task priority field stores an integer value of 1 to 5 to directly reflect the task urgency degree through the numerical size. After the point cloud filtering algorithm removes outliers from the container three-dimensional point cloud data, the contour extraction processing obtains the container height size by identifying the difference between the maximum and minimum values of the Z-axis coordinates in the point cloud data set. The extreme difference values in the X-axis and Y-axis directions correspond to the length and width sizes of the container, respectively. The attitude inclination angle is calculated by fitting the angle between the normal vector of the container top point cloud and the horizontal plane normal vector to obtain the angle parameter reflecting the container placement attitude.

[0037] In the outbound and inbound ratio calculation process, the total number of outbound record rows is counted as the numerator from the task type field of the historical operation record database, and the sum of the outbound record rows and the inbound record rows is the denominator. The ratio result reflects the outbound operation intensity of the warehouse area in the statistical period. The weight dispersion calculation involves two layers of data processing. The first layer extracts the value set of the weight field from the container attribute table to calculate the arithmetic mean as the weight central tendency index. The second layer calculates the square sum of the difference between each weight value and the average value, divides the sample number, and takes the square root to obtain the weight standard deviation reflecting the dispersion degree. The ratio of the standard deviation to the average value eliminates the dimension influence to form the dimensionless dispersion index. The unstacking pressure index uses the number of record rows marked as unstacking in the operation log table as the numerator, and the total number of outbound tasks as the denominator. The ratio quantifies the average number of unstacking operations required for each outbound operation, reflecting the influence of the warehouse layout on the outbound efficiency.

[0038] In a specific embodiment, in step S2, the target weight vector is calculated, including:

[0039] reading a reference weight vector;

[0040] When the warehouse-in and warehouse-out ratio is greater than 0.6, the warehouse-out frequency gain coefficient is calculated, and the warehouse-out frequency reference weight is added to the warehouse-out frequency gain coefficient to obtain the target warehouse-out frequency weight;

[0041] When the weight dispersion is greater than 0.3, the weight balance gain coefficient is calculated, and when the stack pressure index is greater than 0.2, the stack stability gain coefficient is calculated;

[0042] The target warehouse-out frequency weight, the target weight balance weight, the target stack stability weight, and the distance convenience reference weight are normalized to obtain a target weight vector.

[0043] Specifically, the reference weight vector is stored in the weight configuration table as a system initialization parameter, and the reading operation obtains four preset values of the warehouse-out frequency reference weight 0.35, the weight balance reference weight 0.25, the stack stability reference weight 0.25, and the distance convenience reference weight 0.15 by querying the table. The calculation trigger condition of the warehouse-out frequency gain coefficient is that the warehouse-in and warehouse-out ratio value is greater than the threshold 0.6, and the calculation logic is that the difference between the warehouse-in and warehouse-out ratio and the threshold 0.6 is multiplied by the gain amplification coefficient 0.5, and the gain coefficient is added to the warehouse-out frequency reference weight 0.35 to form the target warehouse-out frequency weight. The design of the gain mechanism automatically increases the weight proportion of the warehouse-out frequency factor in the bin evaluation when the warehouse-out operation is frequent. The weight balance gain coefficient and the stack stability gain coefficient use the same condition trigger and calculation mode. When the weight dispersion is greater than 0.3, the weight balance gain coefficient is calculated by multiplying the difference between the weight dispersion and 0.3 by 0.4 and adding it to the weight balance reference weight 0.25. When the stack pressure index is greater than 0.2, the stack stability gain coefficient is calculated by multiplying the difference between the stack pressure index and 0.2 by 0.5 and adding it to the stack stability reference weight 0.25.

[0044] The normalization process corrects the situation that the sum of the four weight components may exceed 1. First, the arithmetic sum of the target warehouse-out frequency weight, the target weight balance weight, the target stack stability weight, and the distance convenience reference weight is calculated as a normalization factor. Then, the four weight components are divided by the normalization factor to obtain the normalized weight values. The normalization operation ensures that the sum of the four weight components is strictly equal to 1, satisfying the normalization constraint condition of the probability distribution. The target weight vector is formed by the four normalized weight components in the fixed order of warehouse-out frequency, weight balance, stack stability, and distance convenience, and is combined into a one-dimensional array structure. This vector is transmitted to the weight adjustment module as the target state parameter of the subsequent weight smoothing transition calculation.

[0045] In a specific embodiment, the difference between the current weight vector and the target weight vector is linearly interpolated by 5 days to obtain a daily adjustment increment, including:

[0046] reading a current weight vector, subtracting each weight component in the target weight vector from a corresponding weight component in the current weight vector to obtain a weight difference value;

[0047] multiplying the weight difference value by 0.2 to obtain a daily adjustment increment;

[0048] according to the adjustment day number n, adding each weight component in the current weight vector by a corresponding daily adjustment increment multiplied by n to obtain each weight component on the day, and combining the weight components on the day to obtain a weight vector on the day.

[0049] Specifically, the current weight vector is read from the weight storage module to include an array structure containing four components of the current outbound frequency weight, the current weight balance weight, the current stacking stability weight, and the current distance convenience weight. The weight difference value is calculated by a component-by-component subtraction operation, that is, the first component of the target weight vector is subtracted from the first component of the current weight vector to obtain the outbound frequency weight difference value, the second component is subtracted to obtain the weight balance weight difference value, the third component is subtracted to obtain the stacking stability weight difference value, and the fourth component is subtracted to obtain the distance convenience weight difference value. The four difference values reflect the direction and amplitude of weight adjustment. The calculation of the daily adjustment increment multiplies the four weight difference values by a fixed coefficient 0.2. The coefficient corresponds to the reciprocal of the 5-day adjustment period, that is, one-fifth of the total adjustment amount completed each day. The outbound frequency daily adjustment increment, the weight balance daily adjustment increment, the stacking stability daily adjustment increment, and the distance convenience daily adjustment increment form an increment vector and are stored in the weight adjustment cache.

[0050] The adjustment day number n is read from the weight adjustment counter to indicate the current day in the 5-day adjustment period, and the value range is an integer from 1 to 5. The calculation of each weight component on the day uses an accumulation formula, that is, the first component of the current weight vector is added to the outbound frequency daily adjustment increment multiplied by n to obtain the outbound frequency weight on the day, and the same accumulation logic is used to calculate the weight balance weight on the day, the stacking stability weight on the day, and the distance convenience weight on the day. The weight vector on the day is formed by combining the four weight components on the day in a fixed order to form a one-dimensional array structure. The vector is updated once a day during the weight adjustment period until the n value reaches 5, and the weight vector on the day converges to the target weight vector. The weight storage module overwrites the current weight vector field with the weight vector on the day to complete a weight state update.

[0051] In a specific embodiment, it further comprises:

[0052] After 7 days of weight adjustment implementation, the total number of reverse stacking operations and the total running distance of the crane are counted, and the baseline reverse stacking number and the baseline running distance before weight adjustment are read;

[0053] The reverse stacking number change rate and the running distance change rate are calculated;

[0054] When the change rate of the number of times of piling is less than an expected piling reduction threshold, a pile stability compensation gain is calculated and the target pile stability weight is corrected;

[0055] When the change rate of the running distance is less than an expected distance reduction threshold and the ratio of the frequency weight to the distance convenience weight is greater than 4, the distance convenience weight is corrected and stored in the weight correction parameter table.

[0056] Specifically, the weight adjustment effect evaluation module starts a data statistics process on the 7th day after the weight adjustment is completed, queries the number of record rows with the timestamp field located within 7 days after the weight adjustment is implemented and the operation type field marked as piling from the operation log database to obtain the total number of piling operations, queries the running track table of all cranes to extract the values of the lateral movement distance field, the longitudinal movement distance field and the vertical movement distance field, sums up the values for each crane and then accumulates the total running distance of all cranes. The reference piling number and the reference running distance are read from the historical reference data table, which records the number of piling operations and the running distance of the crane within the same time span of 7 days before the weight adjustment as the control reference value. The change rate of the number of times of piling is calculated by subtracting the total number of piling operations from the reference piling number and then dividing the result by the reference piling number to obtain a ratio index reflecting the reduction amplitude of the number of times of piling. The change rate of the running distance is calculated by subtracting the total running distance of the crane from the reference running distance and then dividing the result by the reference running distance to obtain a ratio index reflecting the shortening degree of the running distance.

[0057] The calculation trigger condition of the pile stability compensation gain is that the change rate of the number of times of piling is less than the expected piling reduction threshold of 0.15. The calculation logic is to multiply the difference between the expected piling reduction threshold of 0.15 and the actual change rate of the number of times of piling by a compensation coefficient of 0.1 to obtain the value of the pile stability compensation gain. The compensation gain is added to the target pile stability weight to form the corrected target pile stability weight, which is written into the corresponding field of the weight correction parameter table. The trigger condition for correcting the distance convenience weight includes two judgment logics. The first layer judges whether the change rate of the running distance is less than the expected distance reduction threshold of 0.1. The second layer judges whether the ratio of the frequency weight to the current distance convenience weight is greater than the threshold of 4. When both conditions are met, the correction operation is performed, i.e., the current distance convenience weight is added to the fixed compensation gain of 0.05 to obtain the corrected distance convenience weight. The weight correction parameter table stores the corrected target pile stability weight and the corrected distance convenience weight. The table is read in the next round of weight self-adaptive calculation, the correction parameters are added to the newly calculated target weight vector to form a weight iterative optimization closed loop based on the feedback of the actual operation effect.

[0058] Figure 2FIG. 1 is a schematic diagram of the convergence process of the weight feedback correction algorithm in the embodiments of the present application. The diagram shows the change trend of the reduction rate of the number of backstopping and the reduction rate of the running distance in 10 adjustment cycles through a broken line graph. The solid line represents the reduction rate of the number of backstopping, which gradually increases from the initial 0% and reaches 41% in the 7th adjustment cycle, and then tends to be stable and converges. The dashed line represents the reduction rate of the running distance, which increases from 0% to 26% in the 7th cycle, also showing a convergence feature. The expected backstopping reduction threshold of 15% and the expected distance reduction threshold of 10% marked in the diagram are two horizontal reference lines. When the actual reduction rate breaks through these two threshold lines in the 2nd and 3rd cycles, respectively, the system determines that the weight adjustment effect reaches the expected target. The convergence point marked in the 7th cycle indicates that the weight feedback correction mechanism completes the iterative optimization at this time. Thereafter, the growth rate of the reduction rate of the number of backstopping and the reduction rate of the running distance significantly slows down, indicating that the weight configuration has converged to the optimal solution for the current working scenario. The convergence curve in the diagram verifies the effectiveness of the dynamic weight adaptive algorithm in achieving continuous performance improvement and finally stability through feedback correction.

[0059] In a specific embodiment, step S3 comprises:

[0060] Retrieving historical out-of-warehouse records matching the container type, calculating the out-of-warehouse probability by dividing the number of out-of-warehouse times in the last 30 days by the total number of out-of-warehouse times in the last 90 days and multiplying by 3;

[0061] Querying the deviation of the total weight of the containers in the area where the candidate storage location is located from the average weight of the entire warehouse to obtain the weight deviation, and querying the load-bearing margin of the candidate storage location and the spatial distance to the out-of-warehouse channel;

[0062] Multiplying the out-of-warehouse probability by the daily out-of-warehouse frequency weight in the daily weight vector to obtain the out-of-warehouse frequency score item, multiplying the weight deviation obtained by subtracting 1 from the weight balance weight in the daily weight vector to obtain the weight balance score item, multiplying the load-bearing margin by the daily stacking stability weight in the daily weight vector to obtain the stacking stability score item, and multiplying the maximum distance threshold minus the spatial distance by the current distance convenience weight in the daily weight vector to obtain the distance convenience score item;

[0063] Summing up the out-of-warehouse frequency score item, the weight balance score item, the stacking stability score item, and the distance convenience score item to obtain the fitness score.

[0064] Specifically, the outbound probability prediction module matches and queries the historical outbound record database by the container type field, filters all historical record rows with the same container type field value as the container type to be stored, further filters the timestamp field in the filtering result to obtain the number of outbound record statistics in the last 30 days as the short-term outbound frequency, expands the time window to obtain the total number of outbound records in the last 90 days as the long-term outbound baseline, and obtains the outbound frequency ratio of 30 days to 90 days by dividing the short-term outbound frequency by the long-term outbound total frequency. The ratio is multiplied by the amplification coefficient 3 to realize the time scale conversion from the 90-day baseline to the 30-day prediction window to obtain the outbound probability estimation value in the next 30 days.

[0065] The weight deviation query module first reads the area number field of the candidate storage location number from the storage location attribute table, queries all container records matched in the storage data table through the area number association, accumulates the weight field values of these container records to obtain the total weight of the containers in the area where the candidate storage location is located, obtains the average weight of the whole warehouse by querying the weight field of all container records in the storage data table, and obtains the weight deviation index reflecting the deviation degree of the weight distribution in the area by taking the absolute value of the difference between the total weight of the containers in the area where the candidate storage location is located and the average weight of the whole warehouse and dividing the result by the average weight of the whole warehouse.

[0066] The bearing residual amount query reads the design bearing upper limit field and the current carried weight field in the candidate storage location attribute table, subtracts the two to obtain the value of the remaining loadable weight of the candidate storage location, and the spatial distance query reads the three-dimensional coordinates of the center point of the storage location from the candidate storage location attribute table, including the horizontal coordinate, the vertical coordinate and the vertical coordinate. The three-dimensional coordinates of the nearest outbound channel are read from the outbound channel configuration table. The spatial distance value is calculated by taking the square root of the sum of the squares of the differences between the horizontal coordinates, the vertical coordinates and the vertical coordinates of the candidate storage location and the outbound channel.

[0067] The outbound frequency score item calculation multiplies the outbound probability value by the first component of the daily weight vector, i.e. the outbound frequency weight of the day, to obtain the weighted score of this dimension. The weight balance score item calculation adopts the difference between 1 and the weight deviation to represent the weight distribution balance degree, and then multiplies the second component of the daily weight vector, i.e. the weight balance weight of the day. The stacking stability score item directly multiplies the bearing residual amount value by the third component of the daily weight vector, i.e. the stacking stability weight of the day. The distance convenience score item is obtained by subtracting the spatial distance from the maximum distance threshold and dividing the result by the maximum distance threshold to realize the normalization and reverse mapping of the distance, i.e. the smaller the distance, the higher the score. The normalized distance convenience index is multiplied by the fourth component of the daily weight vector, i.e. the current distance convenience weight, to obtain the weighted score of the distance dimension.

[0068] The adaptation degree score is obtained by directly summing the four values of the out-of-warehouse frequency score item, the weight balance score item, the stack stability score item and the distance convenience score item to obtain a comprehensive evaluation index of the candidate storage location. The larger the score value is, the more the candidate storage location meets the container storage requirements under the current weight configuration. The storage location evaluation module repeats the above calculation process to obtain the adaptation degree score of each storage location in the set of all candidate storage locations. The scoring sorting unit sorts all candidate storage locations in descending order of the adaptation degree score, and selects the storage location number at the top of the sorting, i.e., the storage location with the highest adaptation degree score, as the target storage location and outputs it to the dispatch instruction generation module.

[0069] Figure 3 A schematic diagram of the distribution of each score item of the candidate storage location in the embodiment of the present application. The diagram shows the score distribution of five candidate storage locations in the out-of-warehouse frequency score item, the weight balance score item, the stack stability score item and the distance convenience score item. The out-of-warehouse frequency score item of the storage location D is the highest, reaching 0.25, reflecting that the storage location is close to the out-of-warehouse passage and the historical out-of-warehouse record shows that the corresponding container type has a high-frequency out-of-warehouse feature. The weight balance score item of the storage location C is 0.11, indicating that the current total weight of the containers in the area where the storage location is located deviates less from the average weight of the whole warehouse, which is beneficial to the weight distribution balance. The stack stability score item of the storage location E reaches 0.14, indicating that the storage location has sufficient bearing capacity and fewer overhanging containers. Through the weighted sum of the score items in the four dimensions, the storage location D obtains the highest comprehensive adaptation degree score and is selected as the target storage location. The diagram uses column charts with different filling patterns to intuitively compare the advantages and disadvantages of each candidate storage location in different evaluation dimensions, providing a visual basis for storage location allocation decisions.

[0070] In a specific embodiment, step S4 comprises:

[0071] The current position coordinates, working state and queue length of each crane are obtained, and the cranes in a fault state are excluded to obtain a set of available cranes;

[0072] For each crane in the set of available cranes, the spatial distance from the current position to the container source position and the spatial distance from the container source position to the target storage location are calculated, and the total moving distance is obtained by summing. The expected working time is calculated according to the difference in position coordinates and the device moving speed parameter;

[0073] After the total moving distance, the expected working time and the queue length are normalized, respectively, the weighted sum is obtained by weighting the coefficients 0.4, 0.35 and 0.25, and the crane with the highest weighted sum result is selected to generate the dispatch instruction.

[0074] Specifically, the crane status query module reads all crane records from the crane status table of the equipment monitoring database, each record contains a crane number field, a current position coordinate field including three subfields of lateral position coordinate, longitudinal position coordinate and vertical position coordinate, a job status field using enumeration value to identify three states of idle, in execution or fault, a queue length field recording the number of tasks to be executed by the crane currently.

[0075] The available crane screening logic traverses all crane records to determine whether the job status field is equal to the fault identification value, and the crane records with job status as idle or in execution are retained to form an available crane set, and the crane in fault state is excluded from the scheduling candidate range. The spatial distance calculation is performed twice for each crane in the available crane set. The first calculation is to square the difference between the current lateral position coordinate of the crane and the lateral coordinate of the container source position, the square of the difference between the longitudinal position coordinates, and the square of the difference between the vertical position coordinates, and then take the square root of the sum to obtain the spatial distance from the current position of the crane to the container source position. The second calculation is to square the difference between the lateral coordinate of the container source position and the lateral coordinate of the target storage location, the square of the difference between the longitudinal coordinates, and the square of the difference between the vertical coordinates, and then take the square root of the sum to obtain the spatial distance from the container source position to the target storage location. The sum of the two spatial distance values is the total movement distance of the crane to complete the current lifting task.

[0076] The expected job time calculation involves independent time estimation in three directions. The lateral direction time is obtained by taking the absolute value of the difference between the target storage location lateral coordinate and the current lateral position coordinate of the crane, and then dividing by the lateral movement speed parameter read from the equipment parameter table. The longitudinal direction time is obtained by taking the absolute value of the difference between the longitudinal coordinates, and then dividing by the longitudinal movement speed parameter. The vertical direction time is obtained by taking the absolute value of the difference between the vertical coordinates, and then dividing by the vertical movement speed parameter. Since the crane three-dimensional mechanism has partial parallel motion capability, the expected job time takes the maximum value of the three direction times to represent the bottleneck direction that limits the overall job speed. The maximum value is added to the fixed time overhead read from the job parameter table, including container grabbing time and release time, to obtain the complete expected job time value.

[0077] The normalization processing module performs numerical conversion for the total movement distance, the expected job time and the queue length. The total movement distance normalization is obtained by subtracting the total movement distance from the maximum distance threshold read from the configuration table, and then dividing by the maximum distance threshold to obtain the normalization score in the distance dimension. The expected job time normalization is obtained by subtracting the expected job time from the maximum time threshold, and then dividing by the maximum time threshold to obtain the normalization score in the time dimension. The queue length normalization is obtained by subtracting the quotient of the queue length divided by the upper limit of the queue capacity from 1 to obtain the normalization score in the load dimension. The larger the three normalization score values are, the more obvious the scheduling advantage in the corresponding dimension is.

[0078] The dispatch score calculation multiplies the total moving distance normalized score by a weight coefficient 0.4 to obtain a distance dimension weighted score, multiplies the expected job time normalized score by a weight coefficient 0.35 to obtain a time dimension weighted score, multiplies the queue length normalized score by a weight coefficient 0.25 to obtain a load dimension weighted score, and directly sums the three weighted scores to obtain a comprehensive dispatch score value of the crane.

[0079] The dispatch decision module repeats the above calculation process for each crane in the available crane set to obtain a respective dispatch score, the score comparison unit traverses all dispatch score values to identify the maximum value, extracts the crane number corresponding to the maximum dispatch score as the execution crane, and the dispatch instruction generation module encapsulates the execution crane number, container source position three-dimensional coordinates, target storage location three-dimensional coordinates, task priority value and expected completion timestamp to form a dispatch instruction data packet, which is delivered to the control system of the selected crane through the communication interface to trigger an automated lifting operation execution process.

[0080] The above describes a warehouse container automated operation method based on a three-dimensional mechanism in an embodiment of the present application, and the following describes a warehouse container automated operation system based on a three-dimensional mechanism in an embodiment of the present application. One embodiment of the warehouse container automated operation system based on a three-dimensional mechanism in an embodiment of the present application includes:

[0081] The receiving module is configured to receive a job task data packet and container information, and to count the warehouse-in and warehouse-out ratio, weight dispersion and destacking pressure index in the past 7 days as scene characteristic parameters;

[0082] The decomposition module is configured to calculate a target weight vector when any parameter in the scene characteristic parameters exceeds a preset threshold value for 3 consecutive days, to decompose the difference between the current weight vector and the target weight vector into a daily adjustment increment by linear interpolation for 5 days, and to accumulate the daily adjustment increment to the current weight vector day by day to obtain a daily weight vector;

[0083] The weighting module is configured to predict an out-of-warehouse probability according to a container type, to query the weight deviation, load capacity and distance to the out-of-warehouse channel of a candidate storage location, to weight and sum the out-of-warehouse probability, weight deviation, load capacity and distance with the daily weight vector to obtain an adaptation score, and to select a storage location with the highest adaptation score as a target storage location.

[0084] The scheduling module is configured to calculate the three-dimensional moving distance of a crane to a container source position and the target storage location and the job time, to perform scheduling optimization according to the moving distance, job time and queue length, and to select a crane with the highest dispatch score to generate a dispatch instruction.

[0085] The above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can be modified, or some technical features can be replaced by equivalent features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for automated operation of a warehouse container based on a three-dimensional mechanism, characterized by, The method comprises: Step S1, receiving job task data packet and container information, and counting the warehouse-in and warehouse-out ratio, weight dispersion and pile-up pressure index in the past 7 days as scene characteristic parameters; Step S2, when any parameter in the scene characteristic parameters exceeds the preset threshold value continuously for 3 days, a target weight vector is calculated, including: reading a reference weight vector; when the warehouse-in and warehouse-out ratio is greater than 0.6, a warehouse-out frequency gain coefficient is calculated, the warehouse-out frequency reference weight is added to the warehouse-out frequency gain coefficient to obtain a target warehouse-out frequency weight; when the weight dispersion is greater than 0.3, a weight balance gain coefficient is calculated, and when the pile-up pressure index is greater than 0.2, a pile-up stability gain coefficient is calculated; the target warehouse-out frequency weight, the target weight balance weight, the target pile-up stability weight and the distance convenience reference weight are normalized to obtain the target weight vector; The difference between the current weight vector and the target weight vector is linearly interpolated by 5 days to obtain a daily adjustment increment, and the daily adjustment increment is accumulated to the current weight vector day by day to obtain a daily weight vector, including: reading the current weight vector, subtracting the corresponding weight component in the target weight vector from the corresponding weight component in the current weight vector to obtain a weight difference; multiplying the weight difference by 0.2 to obtain a daily adjustment increment; according to the adjustment day number n, adding the corresponding daily adjustment increment multiplied by n to each weight component in the current weight vector to obtain each weight component on the day, and combining the weight components on the day to obtain the daily weight vector; Step S3, according to the container type, the warehouse-out probability is predicted, the weight deviation, the bearing capacity and the distance to the warehouse-out channel of the candidate storage position are queried, the warehouse-out probability, the weight deviation, the bearing capacity and the distance are weighted and summed with the daily weight vector to obtain an adaptation score, and the storage position with the highest adaptation score is selected as the target storage position; Step S4, the three-dimensional moving distance and the operation time of the crane from the container source position to the target storage position are calculated, the moving distance, the operation time and the queue length are optimized, and the crane with the highest scheduling score is selected to generate a scheduling instruction.

2. The method according to claim 1, wherein, The step S1 comprises: Receiving a job task data packet and parsing the task type field, the task area field, the job task field and the task priority field to obtain the task type, the task area number, the container target number and the task priority value; Obtaining the three-dimensional point cloud data of the container and performing contour extraction processing to obtain the three-dimensional size coordinates and the attitude inclination angle of the container; Extracting the total number of warehouse-out tasks and the total number of warehouse-in tasks in the past 7 days from the historical operation record database, calculating the total number of warehouse-out tasks divided by the sum of the total number of warehouse-out tasks and the total number of warehouse-in tasks to obtain the warehouse-in and warehouse-out ratio; Extracting the weight data of all the warehouse-in containers in the past 7 days to calculate the average weight and the weight standard deviation, dividing the weight standard deviation by the average weight to obtain the weight dispersion, and counting the total number of pile-up operations in the past 7 days divided by the total number of warehouse-out tasks to obtain the pile-up pressure index.

3. The method of claim 1, wherein the method further comprises: Further comprising: Counting the total number of pile-up operations and the total running distance of the crane after the weight adjustment is implemented for 7 days, and reading the reference pile-up number and the reference running distance before the weight adjustment; Calculate the change rate of the number of backstopping and the change rate of the running distance; When the change rate of the number of backstopping is less than the expected backstopping reduction threshold, calculate the backstopping stability compensation gain and correct the target backstopping stability weight; When the running distance change rate is less than the expected distance reduction threshold and the ratio of the out-of-warehouse frequency weight to the distance convenience weight is greater than 4, the distance convenience weight is corrected and stored in the weight correction parameter table.

4. The method of claim 1, wherein the method further comprises: The step S3 comprises: Retrieving historical out-of-warehouse records matching the container type, calculating the out-of-warehouse probability by dividing the number of out-of-warehouse times in the last 30 days by the total number of out-of-warehouse times in the last 90 days and multiplying by 3; Query the deviation of the total weight of the containers in the area where the candidate storage location is located from the average weight of the whole warehouse to obtain the weight deviation, and query the load bearing margin of the candidate storage location and the spatial distance to the out-of-warehouse channel; Multiply the out-of-warehouse probability by the out-of-warehouse frequency weight in the daily weight vector to obtain the out-of-warehouse frequency score item, multiply the weight deviation obtained by subtracting 1 from the weight balance weight in the daily weight vector to obtain the weight balance score item, multiply the load bearing margin by the backstopping stability weight in the daily weight vector to obtain the backstopping stability score item, and subtract the spatial distance from the maximum distance threshold, divide by the maximum distance threshold, and multiply by the current distance convenience weight in the daily weight vector to obtain the distance convenience score item; Sum the out-of-warehouse frequency score item, weight balance score item, backstopping stability score item and distance convenience score item to obtain the fitness score.

5. The three-dimensional-mechanism-based warehouse-container-automation method, according to claim 1, wherein, The step S4 comprises: Obtain the current position coordinates, working state and queue length of each crane, and exclude the cranes in fault state to obtain a set of available cranes; For each crane in the set of available cranes, calculate the spatial distance from the current position to the container source position and the spatial distance from the container source position to the target storage location, sum the total moving distance, and calculate the expected working time according to the position coordinate difference and the device moving speed parameter; Sum the total moving distance, expected working time and queue length after normalization processing respectively, and weight sum according to weight coefficients 0.4, 0.35 and 0.25, and select the crane with the highest weighted sum result to generate the scheduling instruction.

6. A warehouse container automation system based on three-dimensional mechanisms, characterized by The three-dimensional mechanism-based warehouse container automation operation method comprises: The receiving module is used for receiving operation task data packets and container information, and calculating the in-out warehouse ratio, weight dispersion and backstopping pressure index in the last 7 days as scene feature parameters; The decomposition module is configured to calculate a target weight vector when any parameter in the scene characteristic parameters exceeds a preset threshold for 3 consecutive days, including: reading a reference weight vector; calculating an out-of-warehouse frequency gain coefficient when the in-out warehouse ratio is greater than 0.6, adding the out-of-warehouse frequency reference weight to the out-of-warehouse frequency gain coefficient to obtain a target out-of-warehouse frequency weight; calculating a weight balance gain coefficient when the weight dispersion is greater than 0.3, and calculating a stack stability gain coefficient when the stack pressure index is greater than 0.2; and performing normalization processing on the target out-of-warehouse frequency weight, the target weight balance weight, the target stack stability weight, and the distance convenience reference weight to obtain the target weight vector; The difference between the current weight vector and the target weight vector is linearly interpolated by 5 days to obtain a daily adjustment increment, and the daily adjustment increment is accumulated to the current weight vector day by day to obtain a current weight vector, including: reading a current weight vector, subtracting the corresponding weight component in the target weight vector from the corresponding weight component in the current weight vector to obtain a weight difference; multiplying the weight difference by 0.2 to obtain a daily adjustment increment; according to the adjustment number n, adding the corresponding daily adjustment increment multiplied by n to each weight component in the current weight vector to obtain each weight component on the current day, and combining the weight components on the current day to obtain a current weight vector; The weighting module is configured to predict an out-of-warehouse probability according to the container type, query the weight deviation, the weight bearing margin, and the distance to the out-of-warehouse channel of the candidate warehouse position, weight-sum the out-of-warehouse probability, the weight deviation, the weight bearing margin, and the distance with the current weight vector to obtain an adaptation score, and select the warehouse position with the highest adaptation score as a target warehouse position; The scheduling module is configured to calculate the three-dimensional moving distance and operation time of the crane from the container source position to the target warehouse position, perform scheduling optimization according to the moving distance, the operation time, and the queue length, and select the crane with the highest scheduling score to generate a scheduling instruction.

7. The system of claim 6, wherein, The scene characteristic parameters include the in-out warehouse ratio, the weight dispersion, and the stack pressure index in the last 7 days, which are calculated by receiving the job task data packet and the container information, including: The task type, the task area number, the container target number, and the task priority value are obtained by receiving the job task data packet and parsing the task type field, the task area field, the job task field, and the task priority field. The three-dimensional point cloud data of the container is obtained and contour extraction processing is performed to obtain the three-dimensional size coordinates and the attitude inclination angle of the container. The in-out warehouse ratio is obtained by calculating the total number of out-of-warehouse tasks divided by the sum of the total number of out-of-warehouse tasks and the total number of in-warehouse tasks in the last 7 days. The weight dispersion is obtained by calculating the weight standard deviation divided by the average weight of all in-warehouse containers in the last 7 days, and the stack pressure index is obtained by calculating the total number of stack operations in the last 7 days divided by the total number of out-of-warehouse tasks.

8. The system of claim 7, wherein, The three-dimensional moving distance and operation time of the crane to the container source position and the target storage position are calculated, scheduling optimization is performed according to the moving distance, operation time and queue length, the crane with the highest scheduling score is selected to generate a scheduling instruction, and the scheduling instruction comprises: obtaining the current position coordinates, operation state and queue length of each crane, and excluding the cranes in a fault state to obtain a set of available cranes; for each crane in the set of available cranes, calculating the spatial distance from the current position to the container source position and the spatial distance from the container source position to the target storage position, summing to obtain the total moving distance, and calculating the expected operation time according to the position coordinate difference and the device moving speed parameter; after the total moving distance, the expected operation time and the queue length are normalized respectively, weighted summation is performed according to the weight coefficients 0.4, 0.35 and 0.25, and the crane with the highest weighted summation result is selected to generate a scheduling instruction.

Images