Tank stacking and material supplementing linkage control method

Abstract

The invention relates to the field of automatic production control, and discloses a tank stacking and material supplementing linkage control method which comprises the following steps that S1, state monitoring is conducted on a stacking area and a material supplementing area through a distributed sensor network, real-time data are obtained, and the real-time data comprise the tank number, the conveying speed and the production takt; s2, the collected real-time data are transmitted to a central control system, a prediction model is used for predicting the material supplementing demand based on historical data and the real-time data, prediction results are obtained, and the prediction results comprise the material supplementing opportunity, the material shortage amount and the material shortage moment; and S3, according to the prediction result, calculating the optimal opportunity and path of material supplementing through an optimal control algorithm, and ensuring that the material supplementing demand in the production process is met in time. According to the method, the intelligent material supplementing system based on fuzzy control and dynamic prediction optimization is introduced, so that the timeliness of material supplementing, the path scheduling efficiency and the system adaptability in the production process are improved.

Description

Technical Field

[0001] This invention relates to the field of automated production control, specifically a method for linkage control of tank stacking and replenishment. Background Technology

[0002] In modern industrial production, especially in continuous canning and stacking scenarios such as food, chemical, and metal packaging, the orderly stacking of cans and the timely replenishment of materials play a crucial role in production line efficiency. With the continuous improvement of automation levels, traditional timed replenishment and manual operation are no longer sufficient to match the frequently changing production rhythm. Improper timing of replenishment can lead to rhythm disruptions or even production line shutdowns, placing significant pressure on production management.

[0003] In existing technologies, some automatic replenishment systems have begun to achieve initial improvements in the automation level of material handling by setting threshold triggering mechanisms and periodic detection methods, combined with fixed path scheduling. These systems are effective in scenarios with stable cycle times and small load fluctuations, reducing manual intervention and improving system consistency. Furthermore, some solutions have attempted to introduce simple fuzzy logic control methods to determine whether replenishment is needed, thereby improving the flexibility of replenishment responses to some extent. However, overall, they are still mainly based on "passive response," lacking the ability to proactively predict replenishment behavior and achieve globally optimal scheduling.

[0004] However, existing technologies generally suffer from several key shortcomings that directly limit their adaptability in complex production environments. First, most systems lack the ability to predict future material consumption trends, essentially operating on a "replenish only when needed" basis, leading to frequent issues of delayed replenishment and response. Second, the selection of replenishment paths is too static; if multiple tasks occur simultaneously, path conflicts and resource contention can easily cause scheduling chaos. Furthermore, rule-based control methods often fall into either "overreacting" or "doing nothing" when faced with minor cycle time disturbances, lacking flexible adjustment capabilities and exhibiting poor adaptability. Moreover, the limited data acquisition dimensions and simple sensor deployment are insufficient to support accurate decision-making under complex scheduling conditions. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a tank palletizing and replenishment linkage control method, which solves the problems of lack of replenishment demand prediction, rigid path scheduling, inflexible control response, and insufficient sensing information in existing technologies.

[0006] To achieve the above objectives, the present invention provides a method for linkage control of tank palletizing and replenishment, comprising the following steps: S1. The status of the palletizing area and the replenishment area is monitored by a distributed sensor network to obtain real-time data, including the number of tanks, conveying speed and production cycle time. S2. The collected real-time data is transmitted to the central control system, and the replenishment demand is predicted using a prediction model based on historical data and real-time data to obtain the prediction results, which include the replenishment timing, the amount of material shortage, and the time of material shortage. S3. Based on the prediction results, calculate the optimal timing and path for material replenishment using the optimal control algorithm to ensure that the material replenishment needs in the production process are met in a timely manner. S4. Based on fuzzy control rules, adjust the material replenishment amount and path according to the material replenishment demand in the production process.

[0007] Preferably, in step S1, the status monitoring of the palletizing area and the replenishment area via a distributed sensor network includes: Multiple sets of sensor nodes deployed in the palletizing and replenishment areas collect data on tank position, stacking height, movement trajectory, conveying speed, and current cycle time. The multiple sets of sensors include laser rangefinders, visual recognition sensors, and speed detection sensors.

[0008] Preferably, the data collected include the tank position, stacking height, movement trajectory, conveying speed, and current cycle time. The position of the tank and the stacking height are obtained using a laser rangefinder sensor; The movement trajectory and conveying path of the tank are obtained through visual recognition sensors; The speed of the conveyor belt is measured using a speed detection sensor.

[0009] Preferably, in step S2, the prediction of replenishment demand using a predictive model based on historical and real-time data includes: Historical data and current real-time data are input into a combined prediction model for calculation by setting a time sliding window. The prediction model includes an integrated structure based on the autoregressive moving average model ARIMA and the long short-term memory network LSTM model. The prediction model output includes the material consumption rate for a future time period, the critical time point for material shortage, and the replenishment trigger window. The tank consumption in the prediction calculation is: In the formula, D t x represents the tank consumption. t For real-time status input; θ represents the historical average state; θ is the model parameter vector; f is the combined prediction function of the ARIMA and LSTM models.

[0010] Preferably, the Long Short-Term Memory (LSTM) network model includes: The input layer is used to receive historical data and real-time data processed by a time sliding window; Long Short-Term Memory (LSTM) network layers are used to capture temporal dependencies in historical data and process data over long periods of time. The output layer provides predictions of replenishment demand, including future material consumption rates and critical shortage times; the model parameter vector includes weights and biases for training the LSTM network, and a learning rate parameter for adjusting the network output accuracy.

[0011] Preferably, in step S3, calculating the optimal timing and path for material replenishment using the optimal control algorithm includes: The central control system constructs an optimal control model with the objectives of minimizing the feeding delay and path cost, and solves the optimal feeding decision based on the feeding time output by the prediction model and the system state variables. The path cost minimization is defined as: In the formula, J is the objective function for the total path cost; c i (t) is the cost function of the i-th path at time t; α i Weighting coefficients are selected for the paths; n is the total number of possible paths; The control decision is obtained by solving an optimization problem that includes path cost functions and state constraints.

[0012] Preferably, minimizing the feeding delay includes: Based on the timing relationship between the replenishment timing and the production process, the time delay of the replenishment operation is calculated and coordinated with other related production processes to ensure that the replenishment operation is synchronized with other production links. The startup and execution of the feeding system can be optimized by adjusting the response speed of the feeding execution equipment, the selection of the transmission path, and the timing of the control signal issuance. By calculating the waiting time during the replenishment process, the scheduling strategy of each part of the replenishment system can be optimized.

[0013] Preferably, in step S4, adjusting the replenishment amount and path based on the replenishment needs during the production process includes: Based on real-time data during the production process, calculate the magnitude of changes in material replenishment and its impact on the production rhythm; Based on the impact derived from the calculation results, the replenishment amount is adjusted using fuzzy control rules, which include the changes in material shortage, replenishment demand, and production cycle time.

[0014] Preferably, the effects derived from the calculation results include: The impact of changes in material replenishment on production rhythm: By analyzing the effects of different material replenishment amounts on the conveying speed and cycle time of the production line, the material replenishment amount can be further adjusted to maintain a stable rhythm in the production process. The impact of changing material replenishment demand trends, real-time assessment of the impact of adjustments to replenishment volume on the prediction of material shortage times and replenishment timing; the impact of changes in production cycle time on replenishment path selection, adjusting the selection of replenishment paths and replenishment methods based on changes in production cycle time.

[0015] This invention provides a method for linkage control of tank stacking and material replenishment. It has the following beneficial effects: 1. This invention constructs a combined prediction model based on ARIMA and LSTM, integrating historical data and real-time status input, to successfully and accurately predict the timing of material replenishment, the amount of material shortage, and the moment of material shortage. Compared to traditional replenishment mechanisms that rely solely on rules or empirical parameter settings, this invention improves prediction accuracy and addresses pain points such as delayed replenishment response and weak material shortage early warning capabilities.

[0016] 2. By employing a multi-objective optimal control algorithm, a complete decision-making framework is constructed by jointly optimizing the cost and delay of the material replenishment path. Compared to the problem of path selection relying on static scheduling rules in existing technologies, this scheme achieves dynamic optimal scheduling for path selection, solving inherent limitations such as path conflicts and slow response during material transportation.

[0017] 3. By combining a fuzzy control mechanism, dynamic adjustments are made based on the changing trends of material replenishment demand and rhythm disturbances, forming a flexible control strategy. This invention differs from existing methods that only use a fixed replenishment rhythm. It can effectively absorb the impact of prediction deviations and rhythm fluctuations in actual operation, solving the problems of insufficient stability and excessive or insufficient replenishment in complex field conditions inherent in traditional solutions.

[0018] 4. Through multi-dimensional data acquisition and distributed deployment of sensor networks, this invention achieves full-coverage monitoring of the key states of the palletizing and replenishment areas. Compared with the previous single-sensor node structure, this solution improves the timeliness and comprehensiveness of data acquisition, and completely solves the scheduling failure caused by multiple sensor blind spots and discontinuous data. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Please see the appendix Figure 1This invention provides a method for linkage control of tank palletizing and replenishment, comprising the following steps: S1. The status of the palletizing area and the replenishment area is monitored through a distributed sensor network to obtain real-time data, including the number of tanks, conveying speed and production cycle time. Real-time monitoring of the palletizing and replenishment areas is a crucial first step. Comprehensive monitoring of these areas through a distributed sensor network allows for the precise collection of critical production data, such as the number of cans, conveyor speed, and production cycle time. This data provides the foundation for subsequent prediction, optimization, and dynamic adjustments, ensuring smooth operation and timely response throughout the entire production process. This type of monitoring effectively enhances the intelligence of the production line, enabling precise replenishment control and production scheduling.

[0022] Typically, the status monitoring of the palletizing and replenishment areas is achieved through sensors deployed at various key locations. Each sensor node transmits data in real time to the central control system via a wireless or wired network. The data collected by these sensors provides the central control system with detailed information about the real-time status of the production process, allowing for rapid adjustments to production strategies.

[0023] A distributed sensor network, combining laser rangefinders, visual recognition sensors, and speed detection sensors, is employed to monitor the status of the palletizing and replenishment areas. These sensor nodes are deployed at different locations on the production line to collect key data in real time. This data includes the number of tanks, conveying speed, and production cycle time.

[0024] In practice, visual recognition sensors and laser rangefinders are used in combination to accurately detect and record the number of cans in the palletizing area. The laser rangefinders acquire the position of the cans within the stack and determine the number of cans and the height of the stack by measuring the distance between them. Meanwhile, the visual recognition sensors use image recognition algorithms to analyze the position of the cans on the conveyor belt in real time, confirming the presence of each can and its movement trajectory.

[0025] For example, through image processing technology, the system can identify the number and location of each can. When the system detects an anomaly in the number of cans at a certain location, the central control system will make corresponding adjustments to ensure the smooth operation of the palletizing area.

[0026] The system can collect real-time speed data of the conveyor belt using speed sensors installed on it. This data provides crucial information for subsequent replenishment timing calculations. During production, conveyor speed plays a decisive role in the selection of replenishment timing and path. Typically, at higher conveyor speeds, the replenishment system needs to respond more efficiently to avoid disrupting the production rhythm.

[0027] Specifically, the speed detection sensor can monitor the conveyor belt's speed in real time through magnetic induction or photoelectric principles. When the conveyor belt speed changes, the system can respond immediately and automatically adjust the feeding amount to ensure that the feeding operation is synchronized with the production speed.

[0028] Production cycle time data reflects the rhythm of the production process and determines the operating speed of the entire production line. In practice, production cycle time is usually calculated using comprehensive monitoring data. Specifically, by calculating the production cycle of each tank in real time, combined with real-time information such as conveyor belt speed and the number of tanks, the current production cycle time can be accurately determined.

[0029] For example, based on the number of tanks and the conveying speed, the system can calculate the production cycle and determine whether production meets the predetermined pace. When the system detects that the production pace deviates from the normal range, the control system will automatically adjust the replenishment strategy to ensure that production does not stagnate due to delayed replenishment.

[0030] The collected real-time data is transmitted to the central control system via a wireless network. To ensure the real-time performance and accuracy of data transmission, a high-bandwidth, low-latency communication protocol is used to guarantee that the system can acquire key data such as the number of tanks, conveying speed, and production cycle time in a timely manner. After receiving this data, the central control system uses it as input, combining it with historical data and predictive models to further analyze and predict replenishment needs.

[0031] At this point, the central control system can not only understand the current production status, but also predict future replenishment needs. Through integration with other modules, the system can issue replenishment signals at appropriate times and automatically select the optimal replenishment path and timing to ensure the smooth operation of the production line.

[0032] Regarding formulas, the predictive calculation formulas for replenishment timing, material shortage amount, and material shortage time can further enrich the technical description. Let historical data be set as X. historical Real-time data is X real-time The model parameter vector is θ. Based on the time sliding window method, combined with ARIMA and LSTM models, the prediction result can be expressed as: In the formula, This represents the predicted replenishment demand over a future time period, including the amount and timing of material shortages; f is the combined prediction function of the ARIMA and LSTM models; θ is the model's parameter vector, which, after training, can accurately describe the relationship between historical and real-time data.

[0033] In this way, the system can accurately predict future replenishment needs based on a combination of real-time and historical data, further optimize replenishment strategies and paths, and ensure that replenishment needs in the production process are met in a timely manner.

[0034] S2. The collected real-time data is transmitted to the central control system. Based on historical and real-time data, a predictive model is used to predict replenishment needs, yielding prediction results including the timing of replenishment, the amount of material shortage, and the exact time of shortage. The central control system receives data collected from the distributed sensor network in real time, combines it with historical data, and uses a predictive model to predict replenishment needs. The purpose of this process is to accurately predict when replenishment is needed, the amount of material to be replenished, and the specific timing of replenishment during the production process, to ensure the smooth operation of the production process and to respond promptly to possible material shortages.

[0035] In previous steps, the status of the palletizing and replenishment areas was monitored in real time using a distributed sensor network, acquiring key data such as the number of cans, conveying speed, and production cycle time. Next, this real-time data will be transmitted to the central control system, where it will be combined with historical data to use predictive models to forecast replenishment needs, ensuring that replenishment operations can be performed in advance, thereby avoiding production line downtime.

[0036] In this embodiment, the collected real-time data, including the number of tanks, conveying speed, and production cycle time, is transmitted to the central control system via efficient data communication. This data transmission can be achieved through wired or wireless means, ensuring timely and stable data delivery.

[0037] Upon receiving this real-time data, the central control system combines it with historical data to use a comprehensive predictive model to forecast replenishment demand. Specifically, the system employs an integrated structure based on an Autoregressive Moving Average (ARIMA) model and a Long Short-Term Memory (LSTM) network model for replenishment demand forecasting. The model combines historical and real-time data by setting a time sliding window, thereby outputting relevant prediction results for replenishment demand.

[0038] First, the central control system, based on historical data trends and combined with real-time data such as the number of tanks, conveying speed, and production cycle time, inputs this information into the ARIMA model for preliminary replenishment demand forecasting. The ARIMA model can perform autoregressive analysis on historical data using time series analysis methods and calculate the material consumption rate over a future period.

[0039] As an alternative, LSTM models can handle data spanning long time periods by capturing long-term dependencies. LSTM effectively avoids the vanishing gradient problem of traditional RNN models when processing long-term series data, thus it can better capture the relationship between historical data and the current production cycle. The output layer of the LSTM network provides predictions of future replenishment needs, including material consumption rates, critical shortage times, and replenishment trigger windows.

[0040] By combining the outputs of ARIMA and LSTM models, the predictive model can provide comprehensive feed demand forecasts for control systems. Specifically, the prediction results include: Replenishment timing: Predict future replenishment times to ensure that replenishment operations are performed at the correct time.

[0041] Material shortage: Predicts the potential material shortage in the future, helping the control system to prepare for material replenishment in advance.

[0042] Material Shortage Time: Predicts the time window when materials will be short of supplies, helping the system determine when to replenish materials.

[0043] During the prediction and calculation process, the tank consumption D t The following formula can be used for calculation: In the formula, D t x represents the tank consumption. t For real-time status input; θ represents the historical average state; θ is the model parameter vector; f is the combined prediction function of the ARIMA and LSTM models.

[0044] In the aforementioned prediction model, the structure and parameter settings of the LSTM network also require special explanation. The input layer of the LSTM network receives historical data and real-time data processed by a time sliding window. This data passes through the network's Long Short-Term Memory layer, capturing the temporal dependencies between data points, and thus generating the prediction output.

[0045] S3. Based on the prediction results, calculate the optimal timing and path for material replenishment using the optimal control algorithm to ensure that the material replenishment needs in the production process are met in a timely manner. After predicting replenishment demand based on real-time and historical data, the system needs to further translate the prediction results into specific control decisions to achieve proactive scheduling and path optimization of the replenishment process. To ensure that the predicted replenishment demand can be met in a timely manner through the optimal path at the appropriate time, this invention introduces a multi-objective optimization-based control strategy into the central control system to calculate the optimal timing and path for replenishment. This process is closely integrated with the prediction module, using optimization algorithms to solve for the prediction output and feeding the optimization results back to the execution module.

[0046] Based on the parameters output by the aforementioned predictive model, such as the timing of material replenishment, the amount of material shortage, and the duration of the shortage, the central control system comprehensively analyzes various state variables and dynamically plans a material replenishment scheduling scheme with the optimization objectives of minimizing replenishment delay and path cost. This control strategy not only alleviates production imbalances caused by sudden material shortages but also improves the overall coordination of the system operation.

[0047] In this embodiment, a joint optimization control model is constructed within the central control system. This model uses parameters such as the material shortage critical time and material consumption rate output by the prediction module as optimization inputs, and inputs the replenishment path cost function and system scheduling state constraints into the optimization solver.

[0048] The path cost minimization is defined as: In the formula, J is the objective function for the total path cost; c i (t) is the cost function of the i-th path at time t; α i Choose a weighting factor for the path; n is the total number of possible paths.

[0049] Path cost function c i (t) can be comprehensively constructed based on the actual site layout, equipment operating status, path length, and traffic congestion level. Specifically, the path cost can include a weighted combination of factors such as travel time, path switching costs, and energy consumption, expressed as: C i (t)=α·d i +β·τ i (t)+γ·e i (t); In the formula, d i C represents the physical length of path i; i (t) is the path cost function; τ i (t) represents the delay time caused by the current traffic occupancy of the route; e i (t) represents the energy consumption required for transportation along the route; α, β, γ are the corresponding weight parameters, which are set according to the scheduling strategy.

[0050] Furthermore, during the optimization process, the system can introduce state variable constraints, including the number of currently available replenishment vehicles, their initial positions, speed ranges, and available power, forming the following state constraint expression: v min ≤v r (t)≤v max E r (t)≥E th ; In the formula, v r (t) represents the speed of the feeding vehicle r at time t; E r (t) represents its remaining charge; E th The minimum safe operating threshold; v min This refers to the minimum speed of the replenishment vehicle; v max This is the maximum speed of the replenishment vehicle.

[0051] Within the aforementioned optimization framework, the system invokes the optimal control algorithm through the central scheduling module to dynamically calculate the minimum path cost and the replenishment start time, and continuously corrects it based on real-time sensor feedback. After the optimization decision is output, the control signal is sent to the replenishment execution device, instructing it to perform the replenishment task according to the optimal path and time, thereby ensuring precise synchronization of the replenishment behavior.

[0052] As a specific example, during the peak production cycle, when the prediction results show that a certain production node will experience a material shortage in 10 minutes, the system will dynamically select the path with the least traffic delay based on the current traffic status, distance and energy consumption of each replenishment path, and issue a replenishment instruction 3 minutes in advance to ensure that the replenishment operation is completed within 1 minute before the material shortage occurs, thus avoiding fluctuations in the system cycle time.

[0053] S4. Based on fuzzy control rules, adjust the material replenishment quantity and path according to the material replenishment demand in the production process; After calculating the optimal timing and path for material replenishment based on the predictive model, it is still necessary to consider the fluctuations and disturbances that may occur during actual production, such as cycle time fluctuations, transportation delays, and differences in equipment response. To cope with these dynamic changes and further improve the adaptability and flexible control capability of the material replenishment system, this invention introduces a fuzzy control mechanism. Based on existing prediction and optimization results, it dynamically adjusts the replenishment amount and path selection according to the real-time changes in replenishment demand during the production process, thereby achieving flexible compensation and control.

[0054] This fuzzy control module serves as a supplementary logic unit to the aforementioned central control system, forming a closed-loop feedback mechanism with the prediction subsystem and the optimization control module. Its inputs originate from the real-time sensing data stream and the output results of the prediction module. Its outputs serve as the material replenishment correction value and the path optimization weight adjustment value, ultimately acting on the execution layer equipment commands to achieve system-level response adjustment.

[0055] The fuzzy control module uses real-time data such as changes in the number of tanks, current conveying speed disturbances, and cycle offset as input variables. It combines the critical time point for material shortage and the material replenishment requirement output by the prediction module to adjust the material replenishment amount and path selection in real time by establishing a fuzzy rule base.

[0056] Generally, to achieve parallel processing of multidimensional information, fuzzy control systems construct multiple input membership functions, including: Material shortage change rate ΔQ: reflects the change in material replenishment demand per unit time; Beat offset ΔT: Indicates the degree of deviation between the actual beat and the planned beat; Conveying speed disturbance ΔV: used to represent the instability in the transportation process; Prediction offset ΔP: the residual error between the predicted value and the actual consumption value.

[0057] The above input variables are transformed into fuzzy linguistic variables through membership functions. Commonly used linguistic sets include "small", "medium", "large", "rapidly increasing", "slowly decreasing", etc. The specific membership functions can be constructed using trigonometric functions or trapezoidal functions.

[0058] As an alternative, the feed quantity adjustment output variable can be defined as ΔL, representing the correction increment of the current feed strategy, and the path selection adjustment variable can be defined as ΔR, used to correct the weight coefficients of each path in the path cost function. The feed adjustment rule can be expressed as: If ΔQ is large, ΔT is large, and ΔV is medium, then ΔL will increase significantly. If ΔP is small and ΔQ is stable, then ΔR remains unchanged; In the formula, the precise values ​​of the output variables ΔL and ΔR are jointly determined by the fuzzy inference mechanism (such as the Mamdani type or Sugeno type) and the subsequent defuzzification method (such as the centroid method).

[0059] Specifically, the fuzzy control module updates the weight factor w in the path cost function in real time based on the execution response time offset of the current path. i This makes the path cost function become: C′ i (t)=(w i +ΔR i )·C i (t); In the formula, C′ i (t) represents the path cost after fuzzy control correction; ΔR i C represents the weight adjustment value for the i-th path output by fuzzy control. i (t) is the path cost function; w i Let be the weight coefficient of the i-th path.

[0060] In one possible implementation, to enhance the dynamic response capability of the fuzzy control system, the central control system can be equipped with a rolling update mechanism. Whenever a new round of real-time data is updated, the fuzzy control rules are re-evaluated to ensure that the material replenishment adjustment mechanism is adaptive.

[0061] In some embodiments, if the system detects that the predicted deviation of the material replenishment amount exceeds a set threshold within two consecutive cycle periods, the fuzzy control module will directly trigger a high-priority material replenishment strategy and temporarily increase the priority of the short path to ensure a fast response.

[0062] Furthermore, to avoid system oscillation caused by excessively frequent adjustments to the material replenishment amount, a hysteresis filtering mechanism can be introduced into the fuzzy control output, i.e., the following correction can be set: ΔL actual =λ·ΔL new +(1-λ)·ΔL prev ; In the formula, ΔL new This represents the original value of the current fuzzy inference output; ΔL prev The output of the previous cycle is λ; λ is the smoothing factor; ΔL actual Adjusted value for actual material replenishment Through the aforementioned technical means, fuzzy control rules can dynamically adjust material replenishment behavior under disturbances such as different production cycles, path blockages, or prediction deviations, effectively connecting the outputs of the aforementioned prediction and optimization algorithms to form a multi-level linkage intelligent control system.

[0063] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for linkage control of tank stacking and material replenishment, characterized in that, Includes the following steps: S1. The status of the palletizing area and the replenishment area is monitored by a distributed sensor network to obtain real-time data, including the number of tanks, conveying speed and production cycle time. S2. The collected real-time data is transmitted to the central control system, and the replenishment demand is predicted using a prediction model based on historical data and real-time data to obtain the prediction results, which include the replenishment timing, the amount of material shortage, and the time of material shortage. S3. Based on the prediction results, calculate the optimal timing and path for material replenishment using the optimal control algorithm to ensure that the material replenishment needs in the production process are met in a timely manner. S4. Based on fuzzy control rules, adjust the material replenishment amount and path according to the material replenishment demand in the production process.

2. The tank stacking and replenishment linkage control method according to claim 1, characterized in that, In step S1, the status monitoring of the palletizing area and the replenishment area through a distributed sensor network includes: Multiple sets of sensor nodes deployed in the palletizing and replenishment areas collect data on tank position, stacking height, movement trajectory, conveying speed, and current cycle time. The multiple sets of sensors include laser rangefinders, visual recognition sensors, and speed detection sensors.

3. The tank stacking and replenishment linkage control method according to claim 2, characterized in that, The data collected include the tank position, stacking height, movement trajectory, conveying speed, and current cycle time. The position of the tank and the stacking height are obtained using a laser rangefinder sensor; The movement trajectory and conveying path of the tank are obtained through visual recognition sensors; The speed of the conveyor belt is measured using a speed detection sensor.

4. The tank stacking and replenishment linkage control method according to claim 1, characterized in that, In step S2, the prediction of replenishment demand using a predictive model based on historical and real-time data includes: Historical data and current real-time data are input into a combined prediction model for calculation by setting a time sliding window. The prediction model includes an integrated structure based on the autoregressive moving average model ARIMA and the long short-term memory network LSTM model. The prediction model output includes the material consumption rate for a future time period, the critical time point for material shortage, and the replenishment trigger window. The tank consumption in the prediction calculation is: In the formula, D t x represents the tank consumption. t For real-time status input; θ represents the historical average state; θ is the model parameter vector; f is the combined prediction function of the ARIMA and LSTM models.

5. The tank stacking and replenishment linkage control method according to claim 4, characterized in that, The Long Short-Term Memory (LSTM) network model includes: The input layer is used to receive historical data and real-time data processed by a time sliding window; Long Short-Term Memory (LSTM) network layers are used to capture temporal dependencies in historical data and process data over long periods of time. The output layer provides predictions of replenishment demand, including future material consumption rates and critical shortage times. The model parameter vector includes weights and biases for training the LSTM network, as well as a learning rate parameter for adjusting the network output accuracy.

6. The tank stacking and replenishment linkage control method according to claim 1, characterized in that, In step S3, calculating the optimal timing and path for material replenishment using the optimal control algorithm includes: The central control system constructs an optimal control model with the objectives of minimizing replenishment delay and path cost, and solves the optimal replenishment decision based on the replenishment time output by the prediction model and the system state variables. The path cost minimization is defined as: In the formula, J is the objective function for the total path cost; c i (t) is the cost function of the i-th path at time t; α i Weighting coefficients are selected for the paths; n is the total number of possible paths; The control decision is obtained by solving an optimization problem that includes path cost functions and state constraints.

7. The tank stacking and replenishment linkage control method according to claim 6, characterized in that, Minimizing the replenishment delay includes: Based on the timing relationship between the replenishment timing and the production process, the time delay of the replenishment operation is calculated and coordinated with other related production processes to ensure that the replenishment operation is synchronized with other production links. The startup and execution of the feeding system can be optimized by adjusting the response speed of the feeding execution equipment, the selection of the transmission path, and the timing of the control signal issuance. By calculating the waiting time during the replenishment process, the scheduling strategy of each part of the replenishment system can be optimized.

8. The tank stacking and replenishment linkage control method according to claim 1, characterized in that, In step S4, adjusting the replenishment amount and path based on the replenishment needs during the production process includes: Based on real-time data during the production process, calculate the magnitude of changes in material replenishment and its impact on the production rhythm; Based on the impact derived from the calculation results, the replenishment amount is adjusted using fuzzy control rules, which include the changes in material shortage, replenishment demand, and production cycle time.

9. The tank stacking and replenishment linkage control method according to claim 8, characterized in that, The effects derived from the calculation results include: The impact of changes in material replenishment on production rhythm: By analyzing the effects of different material replenishment amounts on the conveying speed and cycle time of the production line, the material replenishment amount can be further adjusted to maintain a stable rhythm in the production process. The impact of changing material replenishment demand trends, real-time assessment of the impact of adjustments to replenishment volume on the prediction of material shortage times and replenishment timing; the impact of changes in production cycle time on replenishment path selection, adjusting the selection of replenishment paths and replenishment methods based on changes in production cycle time.

Images