Automatic material conveying method and system for elevator car

Abstract

The invention provides an automatic material conveying method and system for an elevator car, and relates to the technical field of industrial automation and intelligent manufacturing. The automatic material conveying method for the elevator car comprises the steps that the historical working state of the elevator car, the real-time working state of the elevator car and a current task list are obtained, multi-target sorting optimization is conducted, an initial task execution scheme is obtained, the elevator car executes the initial task execution scheme, and the real-time working state and the current task list of the elevator car are updated in real time; and evaluating the initial task allocation scheme in real time to obtain a matching degree, and if the matching degree is smaller than a preset matching degree threshold value, inserting the new task into the initial task execution scheme to generate a latest task execution scheme. According to the automatic material conveying method for the elevator car, the technical problem that in the related technology, static adjustment cannot fundamentally solve resource collaboration, dynamic optimization and system fault tolerance is solved.

Description

An automatic material transfer method and system for elevator cars Technical Field

[0001] This invention relates to the field of industrial automation and intelligent manufacturing technology, specifically to an automatic material transfer method and system for elevator cars. Background Technology

[0002] In modern intelligent manufacturing plants, material handling systems are crucial hubs and "vascular networks" connecting various isolated production stages such as diffusion, etching, thin film deposition, and photolithography. Among these, the elevator car (Lifter), as the core equipment for vertical material transport, directly determines the material flow rate, overall equipment efficiency (OEE), and final production capacity of the entire production line through its scheduling efficiency. However, the currently widely used elevator car material handling solutions based on traditional control logic exhibit numerous technical bottlenecks when facing the highly dynamic, complex, and demanding production environment of high-end manufacturing. These bottlenecks are mainly reflected in the following aspects: First, rigid scheduling strategies and poor adaptability. Existing systems mostly adopt scheduling mechanisms based on fixed priorities (such as carrier ID, task generation time, etc.) or simple round-robin scheduling. These static strategies cannot perceive real-time changes in the production line's status, such as "demand warnings" generated when specific process equipment is about to complete processing, the insertion of sudden high-priority production instructions, or path failures caused by material blockages. Their rigid decision-making model makes it difficult to achieve dynamic optimization oriented towards production cycle time, resulting in the system's inability to flexibly adapt to the needs of modern flexible manufacturing.

[0003] Second, resource coordination efficiency is low, limiting system throughput. In complex transport networks consisting of multiple elevator cars, multiple horizontal conveyor lines, and numerous loading and unloading ports, the coordinated operation between the cars and conveyor lines is crucial. Existing solutions often lack a global resource view and collaborative optimization algorithms, easily leading to the contradictory phenomenon of "equipment idleness and task congestion coexisting." For example, one car may be idle while another car is queuing due to concentrated tasks; or the cycle times of the horizontal conveyor line and the vertical car may be mismatched, forming a transport bottleneck, thereby reducing the overall throughput and resource utilization of the entire material transport system.

[0004] Third, the system has weak fault tolerance and robustness. When a local fault occurs in the system, such as a mechanical failure at a loading / unloading port, an anomaly in the communication network, or a sensor malfunction, traditional control systems can usually only trigger a simple alarm and stop the relevant equipment, relying on manual intervention for troubleshooting and recovery. This passive approach leads to excessively long system recovery times, a wider range of fault impacts, severely drags down overall production efficiency, and may even cause backlogs and damage to work-in-process (WIP), failing to meet the high reliability requirements of 24 / 7 uninterrupted production.

[0005] Fourth, the control system suffers from insufficient real-time performance and significant response latency. High-end manufacturing places extremely stringent demands on the timeliness of material handling; any unnecessary delays can directly impact product yield and production cycles. Traditional centralized or hierarchical control architectures, due to their high computational complexity and communication overhead, exhibit significant system response latency when facing massive concurrent tasks and real-time path planning requests. This makes it difficult to achieve millisecond-level optimization decisions and fails to meet the ultimate pursuit of "real-time perception, real-time analysis, and real-time decision-making" in future intelligent manufacturing.

[0006] To address the aforementioned issues, some improvements have been attempted in existing technologies. For example, scheduling automation is achieved by setting fixed priority rules, which simplifies the control logic to some extent. However, it is essentially still within the scope of static setting and has significant limitations in multi-car collaborative work, dynamic path planning, and coping with complex and ever-changing production environments. It cannot fundamentally solve core issues such as resource coordination, dynamic optimization, and system fault tolerance.

[0007] Therefore, existing technologies need further development. Summary of the Invention

[0008] The purpose of this invention is to overcome the above-mentioned technical deficiencies and provide an automatic material transfer method and system for elevator cars, so as to solve the technical problems of resource coordination, dynamic optimization and system fault tolerance that static setting cannot fundamentally solve in related technologies.

[0009] To achieve the above technical objectives, the present invention adopts the following technical solution: A method for automatic material transfer in an elevator car is provided, comprising: acquiring the historical working status of the elevator car, the real-time working status of the elevator car, and the current task list; the working status includes the current position, current load, running direction, and port availability of the elevator car; performing multi-objective sorting optimization based on the historical working status, the real-time working status, and the current task list to obtain an initial task execution plan; the multi-objective sorting optimization includes a first priority and a second priority, the first priority being workload balancing and the second priority being distance optimization; the elevator car executing the initial task execution plan and updating the real-time working status and the current task list in real time; evaluating the initial task allocation plan in real time to obtain a matching degree, wherein the matching degree is specifically the increase ΔT in the total completion time of the initial task execution plan due to the insertion of a new task; if the matching degree is less than a preset matching degree threshold, inserting the new task into the initial task execution plan to generate a latest task execution plan.

[0010] Furthermore, the automatic material transfer method for elevator cars also includes real-time monitoring of the elevator car equipment status. When a fault is detected, a fault-tolerance process is triggered. The fault-tolerance process includes: dynamically searching for and selecting the optimal backup port from the set of available ports based on preset selection criteria, and rerouting the tasks affected by the fault to the backup port; if no optimal backup port is found, suspending the transfer task queue related to the faulty port, and issuing an alarm notification to maintenance personnel through an audible and visual alarm and / or a manufacturing execution system work order.

[0011] Furthermore, the fault-tolerant process includes automatically triggering a task queuing or exception handling process if no available car is available to receive the task. The exception handling process includes activating backup equipment or notifying manual intervention.

[0012] Furthermore, the multi-objective sorting optimization method includes: sorting the elevator cars in ascending order according to the number of tasks to be executed, and prioritizing the elevator car with the lightest load; if there are multiple elevator cars with similar loads, sorting them in ascending order according to the physical distance between the current position of the elevator car and the starting position of the task, and prioritizing the elevator car with the closest distance.

[0013] Furthermore, the method for obtaining the matching degree includes: defining the current path of the elevator car as a sequence starting from its current position and passing through the start and destination points of all assigned tasks in sequence; when a new task arrives, determining whether the current path of the elevator car has already crossed the start or destination task points in the new task; if not, calculating a new path for the elevator car, the new path being the path formed by inserting the start and destination points of the new task into the unexecuted part of the current path; calculating the total completion time T of the latest task execution plan containing the new path; obtaining the increase ΔT based on the total completion time T of the new path and the total completion time T0 of the initial task execution plan; if ΔT is less than a preset matching degree threshold, inserting the new request into the stopping sequence of the elevator car.

[0014] Furthermore, if the matching degree ΔT of multiple cars is lower than a preset threshold, the new request will be assigned to the car with the smallest ΔT.

[0015] Furthermore, the automatic material transfer method for elevator cars also includes real-time optimization of the latest task execution plan. The optimization methods include: continuous performance evaluation through a variable window period, wherein the window size is dynamically adjusted according to the system load; within each window period, a global optimization strategy based on a genetic algorithm and a real-time adjustment strategy based on reinforcement learning are run in parallel; based on the performance comparison results, better algorithm parameters are automatically selected, and system parameters are adjusted through a smooth migration mechanism.

[0016] Furthermore, optimization methods include a negative correlation between window size and system load.

[0017] An automatic material transfer system for elevator cars is also provided. The system includes: a data acquisition unit for acquiring the historical operating status, real-time operating status, and current task list of the elevator car; the operating status includes the current position, current load, running direction, and port availability of the elevator car; a sorting unit for performing multi-objective sorting optimization based on the historical operating status, real-time operating status, and current task list of the elevator car to obtain an initial task execution plan; the multi-objective sorting optimization includes a first priority and a second priority, where the first priority is workload balancing and the second priority is distance optimization; an execution unit for executing the initial task execution plan in the elevator car and updating the real-time operating status and current task list of the elevator car in real time; an evaluation unit for evaluating the initial task allocation plan in real time and obtaining a matching degree, where the matching degree is specifically the increase ΔT in the total completion time of the initial task execution plan due to the insertion of a new task; and a dynamic adjustment unit for inserting a new task into the initial task execution plan and generating a new task execution plan if the matching degree is less than a preset matching degree threshold.

[0018] A computer-readable storage medium is also provided, on which computer-readable instructions are stored, which, when executed by a processor, implement the steps of the automatic material transfer method for an elevator car or the automatic material transfer system for an elevator car as described above.

[0019] Beneficial effects: 1. The automatic material transfer method for elevator cars of the present invention achieves a significant improvement in material transfer efficiency of more than 50% through the collaborative work of multi-objective optimization algorithm and dynamic scheduling. Among them, the combination of multi-objective sorting optimization based on genetic algorithm and real-time dynamic adjustment based on reinforcement learning effectively solves the rigidity problem of traditional scheduling strategy. At the same time, adaptive load balancing and path optimization based on real-time task matching degree make the equipment utilization rate stable at more than 95%. The millisecond-level high-precision control system and dual-channel verification mechanism ensure the accuracy and real-time performance of the transmission process. Finally, an efficient, reliable and intelligent material transfer solution is constructed, which solves the technical problems of resource coordination, dynamic optimization and system fault tolerance that static setting cannot fundamentally solve in related technologies.

[0020] 2. The automatic material transfer method for elevator cars of the present invention reduces equipment failure recovery time by 80% through a multi-layer fault detection mechanism and intelligent backup port switching strategy, greatly enhances system reliability, and comprehensively improves the overall efficiency of intelligent manufacturing production lines. Attached Figure Description

[0021] Figure 1 is a flowchart of the automatic material transfer method for elevator cars used in an embodiment of the present invention; Figure 2 is a schematic diagram of the automatic material transfer system for elevator cars used in an embodiment of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] According to an embodiment of the present invention, an automatic material transfer method for an elevator car is provided, as shown in Figures 1 and 2. The method includes: S100 acquiring the historical working status of the elevator car, the real-time working status of the elevator car, and the current task list. The working status includes the current position, current load, running direction, and port availability of the elevator car. In this way, by comprehensively and in real-time collecting the historical working status, real-time working status, and current task list of the elevator car, specifically covering key parameters such as the current position, current load, running direction, and port availability of the car, a solid data foundation is laid for subsequent intelligent scheduling and fault-tolerant processing.

[0024] Specifically, the system integrates historical and real-time data to accurately perceive the current status of each elevator car (e.g., elevator car load, real-time position, and direction of travel) and the overall system workload. This provides high-precision input for subsequent multi-objective optimization algorithms (workload balancing and distance-optimal sorting), thereby reducing scheduling delays and decision-making errors and directly contributing to improved material transfer efficiency. At the same time, real-time working status data (e.g., port availability) enables the system to respond quickly to changes. When new tasks arrive, the system can quickly calculate the matching degree or perform fault-tolerant switching, ensuring that the scheduling scheme is always based on the latest system environment, optimizing resource utilization and task completion time.

[0025] S200 performs multi-objective sorting optimization based on the elevator car's historical working status, real-time working status, and current task list to obtain an initial task execution plan. The multi-objective sorting optimization includes a first priority and a second priority. The first priority is workload balancing, and the second priority is optimal distance. In this embodiment, the multi-objective sorting optimization method includes: sorting the elevator cars in ascending order according to the number of tasks to be executed, and prioritizing the elevator car with the lightest load; if there are multiple elevator cars with similar loads, sorting them in ascending order according to the physical distance between the current position of the elevator car and the starting position of the task, and prioritizing the elevator car with the closest distance.

[0026] In this way, multi-objective sorting optimization, which includes first priority (workload balancing) and second priority (distance optimization), achieves a fundamental shift in scheduling decisions from static rules to dynamic intelligence. Through the first priority workload balancing strategy, the system sorts the cars in ascending order of the number of tasks to be executed and prioritizes the car with the lightest load, fundamentally avoiding the "uneven workload" phenomenon common in traditional scheduling. This ensures that the task load of all cars tends to be balanced, directly stabilizing the equipment utilization rate at over 95%. At the same time, by preventing the bottleneck of individual equipment overload, the overall system throughput is significantly improved. Meanwhile, based on load balancing, the second priority distance optimization principle is introduced. When multiple cars have similar loads, the car closest to the task start position is selected first, minimizing the empty running distance and time of the cars. This strategy not only reduced the average waiting time for material transport by more than 40%, but also directly reduced system energy consumption by optimizing the operating path. Furthermore, the two-level priority sorting mechanism constructs an efficient and computationally low decision-making process, enabling the system to complete initial task allocation within milliseconds. This overcomes the technical bottlenecks of traditional optimization algorithms, which are computationally time-consuming and unable to meet real-time scheduling requirements. The multi-objective sorting optimization is not a simple fixed rule, but rather a dynamic calculation based on real-time acquired car status data (position, load, etc.), enabling the system to adapt to changes in the production environment. This provides the core decision-making logic for the intelligent dynamic scheduling engine, ultimately achieving a technical effect of improving the overall material transport efficiency by more than 50%.

[0027] The S300 elevator car executes the initial task execution plan and updates the real-time working status of the elevator car and the current task list in real time. In this way, the real-time status update mechanism constructs a closed-loop control loop of "decision-execution-feedback-re-decision", which enables the system to continuously optimize based on the latest status. When a new task arrives, the matching degree can be calculated immediately based on the current actual path and dynamic insertion can be implemented, thereby reducing the average waiting time by more than 40% and significantly improving the efficiency of a single transport.

[0028] S400 performs real-time evaluation of the initial task allocation scheme to obtain the matching degree. Specifically, the matching degree is the increase ΔT in the total completion time of the initial task execution scheme due to the insertion of a new task. The method for obtaining the matching degree includes: defining the current path of the elevator car as a sequence starting from its current position and sequentially passing through the start and destination points of all allocated tasks; when a new task arrives, determining whether the current path of the elevator car has already crossed the start or destination point of the new task; if not, calculating a new path for the elevator car, which is the path formed by inserting the start and destination points of the new task into the unexecuted portion of the current path; calculating the total completion time T of the latest task execution scheme including the new path; obtaining the increase ΔT based on the total completion time T of the new path and the total completion time T0 of the initial task execution scheme; if ΔT is less than a preset matching degree threshold, inserting the new request into the car's stopping sequence. The matching degree threshold is determined by combining system performance targets (such as average waiting time requirements) with historical operational data statistical analysis and is optimized through experimental testing.

[0029] Preferably, the matching threshold (total delay caused by queue jumping) does not exceed 10 seconds.

[0030] In this way, by introducing a matching degree evaluation mechanism based on completion time increment, dynamic path optimization and real-time resource reallocation are realized in the task scheduling process. By quantitatively calculating the total completion time increment ΔT caused by inserting a new task, the impact of path adjustment on overall efficiency is accurately evaluated, ensuring that each task insertion can be completed within a controllable time cost, avoiding the efficiency loss caused by blind insertion in traditional scheduling. At the same time, the closed-loop decision-making process of "judgment-calculation-comparison-insertion" enables the system to respond to newly arriving tasks in an instant, and make full use of the car's transportation capacity by hitching along the way without affecting the execution of existing tasks.

[0031] If the matching degree is less than the preset matching degree threshold, S500 inserts the new task into the initial task execution plan and generates the latest task execution plan.

[0032] It should be noted that if the matching degree ΔT of multiple cars is lower than the preset threshold, the new request will be assigned to the car with the smallest ΔT.

[0033] In this way, when multiple cars meet the matching conditions, the mechanism of selecting the car with the smallest ΔT to assign tasks can achieve the best global efficiency. By comparing the additional costs incurred by each car due to the insertion of new tasks, it is ensured that each assignment can minimize the impact on the overall system, thereby keeping the equipment utilization rate stable at a high level of over 95%.

[0034] The S600 method for automatic material transfer in elevator cars also includes real-time monitoring of the elevator car equipment status. When a fault is detected, a fault-tolerant process is triggered. The fault-tolerant process includes: S610 dynamically searching for and selecting the optimal backup port from the set of available ports based on preset selection criteria, and rerouting the tasks affected by the fault to the backup port; S620 if the optimal backup port is not found, suspending the transfer task queue related to the faulty port, and issuing an alarm notification to maintenance personnel through an audible and visual alarm and / or a manufacturing execution system work order.

[0035] The S630 fault-tolerant process includes automatically triggering a task queuing or exception handling process if no available car is available to receive the task. The exception handling process includes activating backup equipment or notifying manual intervention.

[0036] It should be noted that the preset selection criteria include recent availability (minimizing material transfer path offset), device compatibility (ensuring that the port supports the current material type), and load balancing (avoiding overload of backup ports).

[0037] In this way, by establishing a systematic fault-tolerant processing mechanism, a highly reliable operational guarantee is provided for the intelligent scheduling system. Based on multiple standards such as spatial distance, equipment compatibility, and load balancing, the intelligent backup port switching mechanism enables the system to reroute tasks within seconds of detecting a fault, minimizing the impact of the fault. Simultaneously, when triggering task suspension and multiple alarms, not only is rapid on-site response achieved through audible and visual alarms, but also digital management of maintenance tasks is achieved through MES (Manufacturing Execution System) work order notifications. This establishes an emergency handling process linking the "site" and the system, significantly improving the standardization and timeliness of fault handling. Finally, the task queuing and anomaly handling mechanism provides the system with a last line of defense. In extreme situations where there are no available cars or backup ports, the system avoids error accumulation and equipment conflicts through orderly suspension and manual intervention requests, demonstrating the safety and completeness of the system design.

[0038] This embodiment organically combines port-level, device-level, and system-level fault tolerance to form a comprehensive protection system covering all fault scenarios. By switching from port-level to low-level device scheduling, a defense-in-depth mechanism is constructed, enabling the system to take appropriate measures when facing faults of different levels, ensuring continuous operational reliability of over 95%.

[0039] In this embodiment, the automatic material transfer method for elevator cars further includes real-time optimization of the latest task execution plan. The optimization method includes: continuous performance evaluation through a variable window period, wherein the window size is dynamically adjusted according to the system load; within each window period, a global optimization strategy based on a genetic algorithm and a real-time adjustment strategy based on reinforcement learning are run in parallel; specifically, the parallel operation mode includes: the system simultaneously executes the genetic algorithm optimization thread and the reinforcement learning decision agent within the window period, the genetic algorithm globally searches for the optimal task plan based on historical data, and the reinforcement learning dynamically adjusts the strategy parameters according to real-time environmental feedback.

[0040] The work of genetic algorithms includes: global optimization of task allocation schemes, generating efficient scheduling sequences through selection, crossover and mutation iterations, with an emphasis on long-term performance improvement.

[0041] The reinforcement learning work includes: learning through real-time status and reward mechanisms, dynamically fine-tuning decisions such as task insertion and sorting, and quickly adapting to system changes.

[0042] In this way, by providing a global optimization benchmark through genetic algorithms and performing real-time corrections through reinforcement learning, a two-layer collaboration of "offline optimization + online adjustment" is formed, ensuring that the solution is both globally efficient and flexible in real time. Using this setup, the overall optimization of the scheduling scheme and its real-time adaptability are achieved, improving the system's average response time by more than 20% and maintaining equipment utilization at a high level of 98%.

[0043] In practice, it is necessary to comprehensively compare the two strategies in terms of multi-dimensional indicators such as response time improvement rate, equipment utilization rate, and energy efficiency.

[0044] Based on the performance comparison results, the system automatically selects the better algorithm parameters and adjusts the system parameters through a smooth migration mechanism.

[0045] It should be noted that the optimization methods include window size being negatively correlated with system load.

[0046] In this way, by running the genetic algorithm and reinforcement learning strategy in parallel, the system can simultaneously consider long-term global optimization and short-term real-time adjustments, achieving synergistic improvements across multiple dimensions such as response time improvement rate, equipment utilization, and energy efficiency. A dynamic adjustment strategy with window size negatively correlated with system load is employed; a smaller window is used for rapid response under high load, while a larger window is used for deep optimization under low load, ensuring optimal performance under various operating conditions. A comprehensive performance evaluation framework is constructed by comparing multiple dimensions such as response time improvement rate, equipment utilization, and energy efficiency, ensuring that system optimization is no longer limited to a single metric but achieves multi-objective synergistic improvement. Through the synergistic effect of dynamic window adjustment, parallel comparison of the two algorithms, and smooth parameter migration, the static optimization algorithm is transformed into an intelligent optimization system with adaptive and self-learning capabilities, providing the core driving force for achieving efficient, stable, and sustainable performance improvements.

[0047] Example 1: This example provides an automatic material transfer method for elevator cars, which specifically includes: 1. Intelligent task allocation 1.1 Candidate car screening First, all currently available elevator cars are screened from the system to ensure that they are in a state that can receive new tasks.

[0048] 1.2 Multi-objective sorting optimization first priority: workload balancing is sorted in ascending order by the number of tasks currently waiting to be executed in the car, and the car with the lightest load is selected first to avoid overloading of a single device.

[0049] Second priority: optimal distance. If multiple cars have similar loads, they are further sorted by their physical distance from the target location, and the car closest to the target location is selected to reduce empty running time and energy consumption.

[0050] 1.3 Task Allocation and Execution: If a suitable car exists, the current task is assigned to that car, and the system status is immediately updated, including: current car status, path transport status, and preparation status of the transport start and destination. 1.4 Recording Task Allocation Log: Real-time updates of car load data trigger downstream control system to execute transport instructions. 1.5 Fault Tolerance Mechanism: If no car is available, the system will automatically trigger task queuing or abnormal handling procedures (such as starting backup equipment or notifying manual intervention). 2. Efficient Fault Tolerance: 2.1 Fault Detection and Diagnosis: Real-time Monitoring of Port Status: The system continuously monitors the operating status of each transmission port (such as communication connection, mechanical status, and sensor data).

[0051] Anomaly detection: When a port status of DOWN (unavailable) is detected, the fault tolerance process is immediately triggered.

[0052] 2.2 Intelligent alternative port switching: Based on spatial distance, device load and task priority, the optimal alternative port is dynamically selected.

[0053] Selection criteria include: recent availability (minimizing material transport path offset), device compatibility (ensuring ports support the current material type), and load balancing (avoiding overload of spare ports).

[0054] Task reassignment: If a suitable backup port is found, the system will automatically reroute the task to that port.

[0055] The recovery operation is also recorded, including: fault timestamp, alternative port ID, and task reassignment details.

[0056] 2.3 Emergency handling when no available port is triggered by the emergency protocol: If the system cannot find an available backup port, the following actions will be taken: suspend related tasks; freeze the current transmission queue to prevent error accumulation.

[0057] Alarm notification: Notify maintenance personnel via audible and visual alarms or MES (Manufacturing Execution System) work orders.

[0058] System self-check: Initiate deep diagnostic mode to locate the root cause (such as network interruption, mechanical jamming).

[0059] 3. Real-time performance optimization: A smart sliding window algorithm is used to achieve dynamic performance optimization.

[0060] The system performs continuous performance evaluation with a variable window period (M tasks). The window size is dynamically adjusted according to the system load: a larger window is used for stable optimization under low load, and a smaller window is used for fast response under high load.

[0061] Preferably, the larger window has 50 tasks; the smaller window has 20 tasks.

[0062] During the optimization process, the system runs two core algorithms in parallel for comparative analysis: the global optimization strategy based on genetic algorithms focuses on long-term performance improvement, while the real-time adjustment strategy based on reinforcement learning focuses on short-term response optimization. At each evaluation, the system recalculates the ideal response times for the past M tasks under both algorithms and comprehensively compares multiple metrics such as response time improvement rate, equipment utilization improvement, and energy efficiency.

[0063] Based on performance comparison results, the system automatically selects superior algorithm parameters and gradually adjusts system parameters through a smooth migration mechanism to avoid system oscillations caused by sudden parameter changes. Simultaneously, a comprehensive performance monitoring system is established to track key indicators such as average latency, equipment utilization, and load balancing in real time, ensuring the system continuously maintains optimal operating conditions. Practical applications show that this optimization scheme reduces the average latency of transmission tasks by more than 40%, stabilizes equipment utilization at over 95%, and can automatically adapt and adjust according to changes in operating conditions.

[0064] This embodiment also provides an automatic material transfer system for elevator cars. The automatic material transfer system for elevator cars includes: a data acquisition unit, which acquires the historical working status, real-time working status, and current task list of the elevator car. The working status includes the current position, current load, running direction, and port availability of the elevator car; a sorting unit, which performs multi-objective sorting optimization based on the historical working status, real-time working status, and current task list of the elevator car to obtain an initial task execution plan. The multi-objective sorting optimization includes a first priority and a second priority, where the first priority is workload balancing and the second priority is distance optimization; an execution unit, which executes the initial task execution plan for the elevator car and updates the real-time working status and current task list of the elevator car in real time; an evaluation unit, which evaluates the initial task allocation plan in real time and obtains a matching degree, wherein the matching degree is specifically the increase ΔT in the total completion time of the initial task execution plan caused by the insertion of a new task; and a dynamic adjustment unit, which inserts a new task into the initial task execution plan and generates a new task execution plan if the matching degree is less than a preset matching degree threshold.

[0065] Example 2: The automatic material transfer system for elevator cars in this example includes: an intelligent dynamic scheduling engine, a fault-tolerant and recovery system, a high-precision control system, and an intelligent monitoring platform.

[0066] The intelligent scheduling engine employs a combination of multi-objective optimization and adaptive load balancing.

[0067] Specifically, the multi-objective optimization algorithm combines genetic algorithms and reinforcement learning to optimize the allocation of transmission tasks in real time; among them, the genetic algorithm is responsible for offline, periodic global optimization, while reinforcement learning handles real-time dynamic adjustments.

[0068] In this embodiment, the genetic algorithm process includes: 1. Population initialization: In the population initialization of the genetic algorithm, each chromosome is explicitly encoded as a set of adjustable rule parameters.

[0069] The scheduling rules specifically include: task urgency weight: priority coefficient for urgent tasks.

[0070] Distance cost weight: The influence coefficient of path length on decision-making.

[0071] Energy cost weight: The influence coefficient of equipment energy consumption on decision-making.

[0072] The encoding uses a direct mapping method, meaning that each gene position on the chromosome directly corresponds to a specific value of a regular parameter (such as [0.7, 0.5, 0.3]).

[0073] The fitness function is as follows: F = w1·U + w2·(1-N_w) + w3·(1-E), where U is the equipment utilization rate, i.e., the proportion of time the car performs tasks per unit time; Nw is the standardized waiting time value, i.e., the value after normalizing the average task waiting time to [0,1]; E is the standardized energy consumption value, i.e., the value after normalizing the total energy consumption of the system to [0,1]; w1, w2, w3 are dynamic weights, which are adjusted in real time according to the system load stage (such as peak / valley) or the focus of the optimization objective.

[0074] It should be noted that the standardization method in this embodiment adopts min-max normalization to eliminate dimensional differences; dynamic adjustment: for example, increasing w2 during peak hours (focusing on reducing waiting time) and increasing w3 during off-peak hours (focusing on energy saving).

[0075] This design enables optimization to balance multiple objectives and adapt to the needs of different operating scenarios.

[0076] The steps of reinforcement learning are as follows: 1. Calculation and dynamic adjustment based on historical data: Learn the pattern of transport requests using long-term historical data, find the most frequently used floor (starting position), and reserve the car accordingly to reduce the average waiting time. 1.1 Data collection: The system records the key information of each transport data, including timestamp, starting point, destination, task completion status, etc.

[0077] 1.2 Divide the day into multiple time windows, such as morning rush hour and evening rush hour.

[0078] 1.3 Calculate the frequency of occurrence of each starting position based on historical data within the time window.

[0079] 1.4 The system automatically selects the appropriate time window model based on the current time; when the car is idle or about to enter an idle state, it is pre-moved to the starting position with the highest starting ratio within the current time window.

[0080] 2. Real-time dynamic adjustment based on recent data: Pay attention to the latest changes in transport demand and make more agile and timely adjustments to the car position to cope with temporary changes in traffic flow.

[0081] 2.1 Data sliding window: The system maintains a fixed buffer queue (such as the last 1000 data transfers).

[0082] 2.2 The system calculates the frequency of each starting position within the sliding window in real time. 2.3 The system makes decisions based on recent data and calculates the proportion of recent starting positions in real time. If it finds that the recent starting proportion of a certain starting position is significantly and continuously higher than the value predicted by the historical model, then the idle car will be dispatched to that starting position.

[0083] 3. Dynamic Based on Real-Time Task Matching: During the execution of the current task by the car, it is determined in real time whether a new task can be executed along the same route, thereby improving the transportation efficiency of a single car and reducing the overall waiting time. 3.1 Path Definition: The current path of the car can be defined as a path starting from the current position c, passing through a series of assigned task points (S1->D1, S2->D2,...), and finally completing all tasks. When a new request (S_new->D_new) occurs, the system needs to verify whether the path is being executed. If not, the generated path is reassigned. If it is being executed, it needs to compare whether the currently executing path has crossed the task points in the new request. If it has not crossed the task points in the new request, the path needs to be recalculated and queued and inserted into the path currently being executed by the car. If it has crossed the task points in the new request, a new path needs to be generated and entered into the waiting state. The cost of reassigning paths in this process is usually time delay.

[0084] 3.2 Matching Degree Calculation: If the starting layer S_new and the destination layer D_new of the new request are both located ahead of the car's current running direction, and the insertion will not cause unacceptable delays in the waiting time of materials inside the car, then the matching degree is high; Matching degree = f(ΔT), where ΔT is the increase in the total completion time of the original task list due to the insertion of the new task. The smaller ΔT is, the higher the matching degree.

[0085] 3.3 For each new request, calculate its matching degree with each running car path in real time; set a matching degree threshold; if there is a car whose matching degree with the new request exceeds the threshold, assign the new task to it and dynamically update the car's stopping sequence; if multiple cars meet the conditions, select the one with the highest matching degree; adaptive load balancing dynamically adjusts the workload of the front and back cars based on real-time monitoring data generated during the interaction, such as car physical status, mechanical load, electrical parameters, and environmental data; furthermore, it also includes fault-tolerant processing, front-end overload detection and diversion, and predictive scheduling, based on historical data analysis such as periodic data, material type, conveying trends, and special events to predict future task requirements and allocate resources in advance.

[0086] The fault tolerance and recovery system adopts a three-level fault detection mechanism. The primary detection includes real-time monitoring of port status; the intermediate detection is used for diagnosing device communication anomalies; the advanced detection assesses the system-level health status; and the recovery system adopts an intelligent recovery strategy, which automatically switches to a backup path or initiates a repair process based on the fault level.

[0087] The high-precision control system optimizes the PLC instruction transmission protocol, reducing control latency to the millisecond level and achieving microsecond-level response; all control instructions adopt a send-acknowledge dual-channel mechanism to ensure reliability; at the same time, the transmission speed is automatically adjusted according to material characteristics and environmental conditions.

[0088] The intelligent monitoring platform collects data from all dimensions, recording key data such as equipment status, task execution, and abnormal events in real time; it provides intuitive system operation status display and early warning functions using a visual dashboard; and it supports remote diagnostic interfaces that are deeply integrated with MES (Manufacturing Execution System) / MCS (Material Control System) through the SECS / GEM (Semiconductor Equipment Communication Standard and General Equipment Model) protocol.

[0089] This embodiment also provides a computer-readable storage medium storing computer-readable instructions. When executed by a processor, the computer-readable instructions implement the steps of the automatic material transfer method for an elevator car or the automatic material transfer system for an elevator car as described above.

[0090] This invention can take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program code. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer-readable storage media include, but are not limited to: novel memories such as phase-change memory / resistive random access memory / magnetic memory / ferroelectric memory (PRAM / RRAM / MRAM / FeRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only optical disc read-only memory (CD-ROM), digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0091] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0092] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments, and will not be repeated here.

[0093] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0094] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0095] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for automatic material transfer in an elevator car, characterized in that, include: The system acquires the historical working status of the elevator car, the real-time working status of the elevator car, and the current task list. The working status includes the current position, current load, running direction, and port availability of the elevator car. Based on the historical working status of the elevator car, the real-time working status of the elevator car, and the current task list, a multi-objective sorting optimization is performed to obtain an initial task execution plan. The multi-objective sorting optimization includes a first priority and a second priority. The first priority is workload balancing, and the second priority is distance optimization. The elevator car executes the initial task execution plan and updates the real-time working status and current task list of the elevator car in real time. The initial task allocation scheme is evaluated in real time to obtain the matching degree, wherein the matching degree is specifically the increase ΔT in the total completion time of the initial task execution scheme caused by the insertion of a new task; if the matching degree is less than a preset matching degree threshold, the new task is inserted into the initial task execution scheme to generate the latest task execution scheme.

2. The automatic material transfer method for an elevator car according to claim 1, characterized in that, The automatic material transfer method for elevator cars further includes real-time monitoring of the elevator car equipment status. When a fault is detected, a fault-tolerance process is triggered. The fault-tolerance process includes: dynamically searching for and selecting the optimal backup port from the set of available ports based on preset selection criteria, and rerouting the tasks affected by the fault to the backup port; if the optimal backup port is not found, suspending the transmission task queue related to the faulty port, and issuing an alarm notification to maintenance personnel through an audible and visual alarm and / or a manufacturing execution system work order.

3. The automatic material transfer method for an elevator car according to claim 2, characterized in that, The fault tolerance process includes automatically triggering a task queuing or anomaly handling process if no available car is available to receive the task. The anomaly handling process includes activating backup equipment or notifying manual intervention.

4. The automatic material transfer method for an elevator car according to claim 1, characterized in that, The multi-objective sorting optimization method includes: sorting the elevator cars in ascending order according to the number of tasks to be executed, and prioritizing the elevator car with the lightest load; if there are multiple elevator cars with similar loads, sorting them in ascending order according to the physical distance between the current position of the elevator car and the starting position of the task, and prioritizing the elevator car with the closest distance.

5. The automatic material transfer method for an elevator car according to claim 1, characterized in that, The method for obtaining the matching degree includes: defining the current path of the elevator car as a sequence starting from its current position and passing through the start and destination points of all assigned tasks in sequence; when the new task arrives, determining whether the current path of the elevator car has already crossed the start or destination task points of the new task; if not, calculating a new path for the elevator car, the new path being the path formed by inserting the start and destination points of the new task into the unexecuted part of the current path; calculating the total completion time T of the latest task execution plan including the new path; obtaining the increase ΔT based on the total completion time T of the new path and the total completion time T0 of the initial task execution plan; if the ΔT is less than a preset matching degree threshold, inserting the new request into the stopping sequence of the elevator car.

6. The automatic material transfer method for an elevator car according to claim 5, characterized in that, If the matching degree ΔT of multiple cars is lower than the preset threshold, the new request will be assigned to the car with the smallest ΔT.

7. The automatic material transfer method for an elevator car according to claim 1, characterized in that, The automatic material transfer method for elevator cars further includes real-time optimization of the latest task execution plan. The optimization method includes: continuous performance evaluation through a variable window period, wherein the window size is dynamically adjusted according to the system load; within each window period, a global optimization strategy based on a genetic algorithm and a real-time adjustment strategy based on reinforcement learning are run in parallel; based on the performance comparison results, better algorithm parameters are automatically selected, and system parameters are adjusted through a smooth migration mechanism.

8. The automatic material transfer method for an elevator car according to claim 7, characterized in that, The optimization method includes the window size being negatively correlated with system load.

9. An automatic material transfer system for an elevator car, characterized in that, The automatic material transfer system for elevator cars includes: a data acquisition unit, which acquires the historical working status, real-time working status, and current task list of the elevator car, wherein the working status includes the current position, current load, running direction, and port availability of the elevator car; a sorting unit, which performs multi-objective sorting optimization based on the historical working status, real-time working status, and current task list of the elevator car to obtain an initial task execution plan, wherein the multi-objective sorting optimization includes a first priority and a second priority, wherein the first priority is workload balancing and the second priority is distance optimization; an execution unit, which executes the initial task execution plan for the elevator car and updates the real-time working status and current task list of the elevator car in real time; an evaluation unit, which evaluates the initial task allocation plan in real time and obtains a matching degree, wherein the matching degree is specifically the increase ΔT in the total completion time of the initial task execution plan due to the insertion of a new task; and a dynamic adjustment unit, which inserts the new task into the initial task execution plan and generates a latest task execution plan if the matching degree is less than a preset matching degree threshold.

10. A computer-readable storage medium storing computer-readable instructions thereon, characterized in that, When the computer-readable instructions are executed by a processor, they implement the steps of any of the automatic material transfer methods for an elevator car as claimed in any one of claims 1-8, or the automatic material transfer system for an elevator car as claimed in claim 9.

Images