Automatic warehouse returning system and method for pneumatic logistics conveying bottles

Abstract

The invention discloses an automatic warehouse returning system and method for pneumatic logistics conveying bottles, and relates to the technical field of automatic control. The position and state information of a transmission bottle is monitored in real time through a sensing system, a recovery task is automatically triggered when a no-load to-be-recovered state is detected, and then a central intelligent scheduling system plans an optimal transmission path by using a built-in algorithm and issues a path instruction to drive a pneumatic logistics network execution mechanism to guide driving; and the network control system drives executing mechanisms such as a pipeline valve and a fan to precisely regulate and control the conveying direction, and finally, after a designated recycling station is reached, physical classification and arrangement are completed through the automatic sorting device, full-process automatic management and control are achieved, and the intelligent level of the hospital pneumatic logistics conveying system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automation control technology, in particular to a pneumatic logistics conveying bottle automatic return system and method. BACKGROUND

[0002] In the operation environment of large hospitals, efficient transmission of medical supplies (such as medicines, blood samples, medical files, medical equipment, etc.) is an important link to ensure the smooth progress of daily diagnosis and treatment activities. The pneumatic logistics transmission system (PTS) uses compressed air to drive and transmit goods in a closed pipeline network, which can realize cross-floor and cross-building delivery of goods. Because of its fast transmission speed and high efficiency, it is applied in hospitals.

[0003] However, with the advancement of hospital informatization construction and the transformation of operation management to refinement, the traditional pneumatic logistics transmission system has deficiencies in the management and recycling of transmission bottles, which affects the overall operation efficiency of the system. As the carrier and circulation unit of the system, the management level of the transmission bottle directly affects the service capacity and operation cost of the system. At present, the transmission bottle recycling method mainly relies on manual return and target-oriented transmission bottle and other semi-automatic methods. Medical staff needs to manually put the empty bottles into the recycling pipeline, or classify them by color, barcode and other physical marks. These traditional methods have low efficiency (recycling delay, resource waste), high error rate (classification / placement errors causing asset out of control), extensive management (unable to track individuals, decision lag), rigid scheduling (lack of dynamic perception prone to congestion), safety hazards (manual contact with contaminated bottles), and high costs (complex mechanical device maintenance), which restrict the intelligent level of the pneumatic logistics transmission system and the improvement of the optimization allocation capability of medical resources. SUMMARY

[0004] The purpose of the present application is to provide a pneumatic logistics conveying bottle automatic return system and method, thereby realizing a full-process automation and intelligent closed-loop management solution from "task completion" to "ready for use", and improving the intelligent level of the hospital pneumatic logistics transmission system and the optimization allocation capability of medical resources.

[0005] To achieve the above purpose, the present application provides the following technical solutions.

[0006] In a first aspect, the present application provides a pneumatic logistics conveying bottle automatic return system, comprising: a plurality of intelligent transmission bottles, each of which is integrated with an Internet of Things chip storing its unique identification information; A perception system, composed of multiple wireless network readers deployed in the pneumatic logistics transportation network, is used to automatically identify and locate, and read the information of the Internet of Things chip of the intelligent transportation bottle in real time as it passes by; A central intelligent scheduling system, in communication with the perception system, is used to receive and process the data uploaded by the wireless network readers, and is internally provided with an intelligent scheduling algorithm that can dynamically plan the optimal transportation path for the intelligent transportation bottle in the empty state according to real-time data; A network control system, connected to the central intelligent scheduling system, is used to receive the optimal transportation path instructions and drive the actuators in the pneumatic logistics transportation network to guide the intelligent transportation bottle to travel along the path; At least one automatic sorting device is deployed at a designated recycling site to automatically physically sort and place the arriving intelligent transportation bottles in the corresponding recycling bins according to the instructions of the central intelligent scheduling system or preset rules.

[0007] Optionally, the intelligent scheduling algorithm running in the central intelligent scheduling system is a heuristic algorithm based on multi-objective weighted scoring.

[0008] Optionally, the heuristic algorithm quantifies and comprehensively evaluates at least the following dimensions of characteristics when making decisions: path cost, task priority, network congestion degree, and energy consumption under resource constraints.

[0009] Optionally, the task priority is dynamically calculated based on the static task attributes associated with the intelligent transportation bottle and the waiting time after it enters the recycling state.

[0010] Optionally, the central intelligent scheduling system can dynamically adjust the weight coefficients used to calculate the multi-objective weighted score according to preset strategies, such as time-of-day strategies or strategies based on the current total load of the system, to address congestion phenomena.

[0011] Optionally, the Internet of Things chip is a passive ultra-high frequency radio frequency identification (UHF RFID) anti-metal flexible tag used to dynamically record the load state and cleanliness state of the intelligent transportation bottle.

[0012] Optionally, the automatic sorting device includes a wireless radio frequency identification reader, a main conveyor belt, and a sorting actuator driven by an automated control system. The wireless radio frequency identification reader is installed at the entrance of the main conveyor belt to read the information of the Internet of Things chip on the intelligent transportation bottle.

[0013] Optionally, the automatic sorting device is also integrated with a visual detection unit of high-efficiency recognition technology to detect the physical integrity of the intelligent transportation bottle before sorting.

[0014] In a second aspect, the present application provides an automatic return-to-stock method for pneumatic logistics transfer bottles, comprising the following steps: S1: state sensing, through the sensing system, real-time monitoring of the position information and state information of each intelligent transfer bottle; S2: task triggering, when it is detected that the intelligent transfer bottle completes the transfer task and enters the empty state for recycling, the recycling task is automatically triggered; S3: scheduling decision, the central intelligent scheduling system executes the built-in intelligent scheduling algorithm to calculate an optimal transfer path for the recycling task; S4: path execution, the central intelligent scheduling system issues the optimal transfer path instruction to the network control system, which drives the execution mechanism in the pneumatic logistics transfer network to guide the intelligent transfer bottle to travel along the path; S5: automatic sorting, when the intelligent transfer bottle reaches the designated recycling station, the automatic sorting device automatically classifies and arranges it physically; S6: closed loop confirmation, after the intelligent transfer bottle is sorted, the central intelligent scheduling system confirms that the recycling task is completed and updates its state information.

[0015] Optionally, in the scheduling decision step, the intelligent scheduling algorithm makes decisions in the following way: Calculate multiple candidate paths through path planning algorithm; Quantitative evaluation of multi-dimensional features of each candidate path, including path cost, task priority, network congestion degree and energy consumption under resource constraints; Calculate and compare the comprehensive scores of each candidate path through a multi-objective weighted scoring function, and select the path with the highest score as the optimal transfer path.

[0016] The beneficial effects of the present application are: (1) Improved operational efficiency: through automatic closed-loop management, the involvement of manual in the recycling link is reduced, allowing medical staff to focus more on core diagnosis and treatment business; intelligent scheduling improves the transfer efficiency and reuse speed of the transfer bottle, effectively alleviating the shortage of bottles caused by delayed recycling, and improving the service capacity and response speed of the logistics system.

[0017] (2) Improved management precision: based on individualized and real-time tracking of each transfer bottle, the present application realizes the transition from extensive management to "one bottle one file" fine management, and the system can build a full life cycle digital file for each bottle, supporting data-driven decisions such as accurate inventory, predictive maintenance, and department inventory optimization.

[0018] (3) System robustness and intelligence improvement: The built-in intelligent scheduling algorithm enables the system to have strong decision-making ability. The system can perceive and avoid network congestion in real time, dynamically respond to hardware failures (such as temporary closure of the recycling center), and has good adaptability and fault tolerance, improving the availability and reliability of the logistics system in complex and unexpected situations.

[0019] (4) Security and cost-effectiveness improvement: Automated processes reduce personnel and potential contamination transmission bottle contact, helping to reduce the risk of hospital cross-infection and improve biosecurity; at the same time, by reducing dependence on semi-automatic mechanical devices and optimizing energy consumption, combined with predictive maintenance to extend equipment life, the present application has good cost-effectiveness in long-term operation. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0021] Figure 1 An overall architecture diagram of the gas logistics transfer bottle automatic recycling system is provided for an embodiment of the present application.

[0022] Figure 2 A structural schematic diagram of the intelligent transfer bottle is provided for an embodiment of the present application.

[0023] Figure 3 A structural schematic diagram of the automatic sorting device is provided for an embodiment of the present application.

[0024] Figure 4 A gas logistics transfer bottle automatic recycling method flowchart is provided for an embodiment of the present application.

[0025] Figure 5 A scheduling decision flowchart is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0027] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0028] Figure 1 This invention provides an overall architecture diagram of an automated pneumatic tube return system for pneumatic tube transport bottles, operating within a typical hospital pneumatic tube network environment. This physical world primarily comprises multiple departmental workstations, multiple intelligent transport bottles, a complex transport network consisting of pipes, commutators, fans, etc., and an automated sorting device. The core system of this invention is built upon this physical network, forming a full-stack solution covering information perception, intelligent decision-making, and closed-loop control, specifically including the following key components: The intelligent transmission bottle is the basic unit for achieving comprehensive perception and the source of information collection in this system. For example... Figure 2 As shown, the smart transmission bottle consists of an IoT chip, a bottle body, and a cap. The bottle body and cap serve as the physical carrier, typically injection-molded from medical-grade high-strength polycarbonate (PC) or similar materials to ensure durability, impact resistance, and biocompatibility. The bottle body is designed with standardized interfaces to accommodate different types of workstation transceiver devices. The IoT chip is the core component for achieving "intelligence." This chip is a passive ultra-high frequency (UHF) radio frequency identification (RFID) anti-metal flexible tag. The selection of UHF RFID technology is based on the following comprehensive considerations: Reading distance and speed: UHF bands (such as 860-960 MHz) provide reading distances of several meters and reliable identification of high-speed moving targets, which is crucial for transport bottles moving at speeds of several meters per second in pipes, while low-frequency (LF) and high-frequency (HF) RFID cannot meet this requirement. Anti-interference: The "anti-metal" design of the tag can effectively avoid the shielding and interference of RFID signals by metal instruments or liquids containing metal ions (such as saline) that may be loaded in the bottle. Physical form: "Flexible labels" can better fit the curved surface of the bottle or cap, simplifying the integration process; Cost and passive characteristics: Passive tags do not require built-in batteries and obtain energy through electromagnetic waves emitted by the reader. They have an extremely long lifespan and low unit cost, making them suitable for large-scale deployment.

[0029] In addition, the storage area of the Internet of Things chip is logically divided into TID area, EPC area and User area to carry different types of data. Among them, the TID area stores the globally unique ID number written by the chip manufacturer, which is used as the unique identification at the physical level; the EPC area is burned by the hospital asset management system, and a globally unique unique identification information (for example, asset number P20230801) is given to the transmission bottle, which is used for identification at the business level; the User area is a repeatable read-write area, which is the key to realize dynamic state tracking, and is used to record the real-time state information of the bottle, such as **[loading state]** (for example, binary "01" represents empty load, "10" represents load), clean state ("cleaned" / "to be cleaned"), current associated task ID, historical use frequency, last cleaning and disinfection timestamp, etc.

[0030] In order to maximize the success rate of signal reading and effectively protect the chip, the Internet of Things chip is preferably embedded or in the form of epoxy resin potting, firmly fixed in the reserved groove inside the bottle cap or the non-metallic area at the bottom of the bottle body. Such layout can effectively avoid impact and wear during transmission.

[0031] The perception system is the data acquisition network of the system, which is composed of multiple wireless network readers deployed in the air logistics transmission network. Its core function is to realize the automatic identification and positioning of all intelligent transmission bottles in the network, so as to realize real-time and accurate collection of dynamic data.

[0032] In this embodiment, the wireless network reader adopts a fixed UHF RFID reader, and each reader can connect 1 to 4 directional or omnidirectional antennas according to the coverage range requirements. The deployment position of the wireless network reader is carefully planned to realize full coverage of the transmission network and accurate judgment of key business actions. Typical deployment positions include entrances and exits of workstations, key branch points of pipeline networks, and special areas such as recycling centers and cleaning and disinfection areas.

[0033] The reader antenna deployed at the entrance and exit of the workstation is used to judge the "origin" and "final arrival" of the transmission bottle, and the RFID data read by it is fused with the sensor signals in the station. For example, when a bottle is put into the sending port, the microswitch trigger signal is associated with the bottle ID read by the RFID reader, and the system can accurately determine that "bottle P001 starts the task at T1 in A station", which improves the accuracy of event judgment.

[0034] Antennas are deployed at the front and back of each pipeline diverter at the key branch points of the pipeline network, forming a logical "entry-verification-exit" monitoring door. This not only can track the path of the bottle in real time, but also can verify whether the diverter has acted correctly according to the instruction by comparing the scheduling instruction with the actual path, so as to realize closed-loop monitoring of the physical execution layer state.

[0035] At the entrance and exit of special areas such as recycling centers, cleaning and disinfection areas, read-write devices are deployed to accurately record the time when the bottles enter and leave, thereby automatically updating their status (e.g. "to be sorted", "cleaning", "in stock").

[0036] All wireless network read-write devices are connected to the hospital's local area network through industrial Ethernet, and to simplify field wiring and power supply, the technology supporting PoE (Power over Ethernet) power supply is preferred. The wireless network read-write device will collect the raw tag data (including but not limited to: EPC code, timestamp, signal strength indication RSSI, antenna number reading the tag) and upload it to the central intelligent scheduling system in real time and uninterrupted through TCP / IP protocol.

[0037] The central intelligent scheduling system is the decision-making core of the system, responsible for data processing, intelligent decision-making, and task assignment. It is usually deployed in software form on the hospital's local server or cloud platform, and its core software modules include data access and preprocessing modules, digital twin and state management modules, intelligent scheduling algorithm engines, task assignment and monitoring modules, and human-computer interaction and data visualization modules. Among them, the data access and preprocessing module, as the entrance of the data lake, is responsible for receiving data streams from all readers. This module has built-in data cleaning algorithms, such as a de-duplication algorithm based on a sliding window or RSSI threshold to filter out repeated readings of a single bottle at the same location, and a Kalman filter algorithm to smooth RSSI values. After cleaning and parsing, the original tag data is converted into structured location events with accurate spatio-temporal information. The digital twin and state management module maintains a real-time digital twin model in memory that corresponds to the physical logistics network. This model is a dynamic, multi-dimensional data structure that not only contains the network's static topological information (such as sites, pipe lengths, and connection relationships), but also synchronously and dynamically updates the state of each intelligent transport bottle (such as accurate location, moving speed, and task status), as well as the dynamic attributes of each pipe, such as congestion index (which can be dynamically calculated based on parameters such as the number of bottles passing through per unit time and the current bottle density in the pipe). When a bottle is taken out of the delivery port by medical staff at a work site (this event is determined by fusing site sensor and RFID data), the module immediately updates its state to "empty and ready for recycling," triggering the subsequent scheduling process. The intelligent scheduling algorithm engine is the decision-making core of the system. When the state management module detects a bottle in the "empty and ready for recycling" state, it immediately invokes this engine, which runs the heuristic scheduling strategy based on multi-objective weighted scoring detailed below. The core advantage of this strategy is its flexibility and adaptability, allowing it to dynamically adjust decision-making behavior based on different optimization objectives (such as time priority, energy consumption priority, or congestion avoidance priority). The task assignment and monitoring module, as a bridge connecting the decision-making layer and the control layer, converts the abstract path instructions generated by the algorithm engine (such as "P001: SiteA→NodeJ1→NodeJ3→RecycleCenterB") into communication protocol formats (such as Modbus TCP / IP, OPC UA, or other private protocols) that meet industrial control standards, and then accurately sends them to the network control system. At the same time, it continuously subscribes to subsequent location events of the bottle, forming a monitoring closed loop to verify whether the task is strictly following the predetermined path and time, and triggering an exception handling mechanism when deviations occur (such as the bottle not arriving at the next node at the expected time).The human-computer interaction and data visualization module provides a web-based or client-based graphical user interface (GUI) that allows system administrators to monitor the entire logistics network in real time from a global perspective (such as network congestion heat map, number of bottles at each site), query the historical trajectory and current state of any bottle, manually adjust the weight parameters of the scheduling strategy (for example, increase the energy-saving weight at night), handle abnormal situation alarms reported by the system (such as persistent congestion of a pipeline, offline of a device), and generate various statistical reports for operational analysis.

[0038] The network control system is a distributed hardware control network composed of PLC (Programmable Logic Controller) or DCS (Distributed Control System), which is the bridge connecting the information world (central scheduling system) and the physical world (pipeline network). It directly controls all physical actuators in the entire pipeline network through strong or weak electrical signals, such as the start and stop of the fan, the switching of the three-way / four-way diverter, and the baffle of the station transceiver device; it receives high-level logical instructions from the central intelligent scheduling system (for example, at T1 time, switch the diverter at J1 node to the direction leading to J3), and translates them into specific, millisecond-level I / O level signals or field bus instructions to drive the corresponding physical devices to complete the action with high reliability and real-time performance.

[0039] The automatic sorting device is the key execution end for realizing the automation of the physical end of the recycling process, and is usually deployed in the recycling center.

[0040] From Figure 3 As can be seen, the entrance guiding section physically decelerates the transport bottles arriving from the main pipeline at high speed and smoothly guides them onto the main conveyor belt. The main conveyor belt entrance is also equipped with an RFID reader antenna for final confirmation of the identity information of the arriving bottles before sorting, achieving data verification.

[0041] To increase the robustness and functional dimension of the system, a visual detection unit (such as an industrial camera and a structured light source) with high-efficiency recognition technology can be added above the main conveyor belt. When the bottle passes through, it takes high-speed photos and analyzes them by the background image processing algorithm. The functions that this unit can achieve include: (1) integrity detection: identifying whether the bottle body has obvious damage or cracks, (2) cleanliness detection: identifying whether the bottle surface has abnormal stains, (3) redundant identification: in the extreme case of accidental failure of the RFID tag, the pre-reserved QR code or color mark on the bottle body can be used as a backup identification means, and the detection result can be used as one of the bases for sorting (such as damaged bottles entering the repair box).

[0042] A plurality of groups of recycling boxes are arranged at both sides or ends of the main conveying belt. Each recycling box corresponds to a classification (such as "to be cleaned", "cleaned and available", "to be repaired", etc.) and is provided with a sorting actuator driven by an automatic control system. The central intelligent scheduling system or the embedded controller locally embedded in the device can accurately activate the corresponding sorting actuator according to the database information associated with the ID of the bottle (such as the type of the bottle, the cleaning state) and the optional visual detection result. According to the requirements for sorting speed and cost, the sorting actuator can have various implementation manners. For example, the pneumatic push rod has the advantages of fast response speed, simple structure and low cost, and is suitable for high-speed sorting scenes; and the swing arm / turning plate mechanism driven by a small servo motor or a stepping motor can provide more flexible action and more accurate positioning control.

[0043] In order to realize unattended operation, a sensor such as a weighing sensor, an infrared reflection sensor or an ultrasonic level meter can be arranged at the bottom or upper portion of each recycling box. When the number or height of the bottles in the box reaches a preset threshold, the system will automatically send an alarm information to the central system to prompt the logistics personnel to replace the recycling box in time.

[0044] The method flow provided by the embodiment of the application is as shown in Figure 4 The core of the method flow is an intelligent scheduling algorithm periodically or event-triggered executed by the central intelligent scheduling system. The following will describe the key steps in the algorithm and the method flow in detail.

[0045] S1: state perception. After the system is started, the static topological data of the logistics network is first loaded from the configuration database, including the physical length, diameter, material (affecting the friction coefficient) of all sites and pipes, the position and switching time of the reverser, etc., so as to construct the basic framework of the digital twin model in the memory. Subsequently, the system enters a 7*24 hour real-time monitoring state, continuously collects and updates the position information and state information of each intelligent conveying bottle through the perception system, and dynamically maintains the digital twin model of the entire network, so that it becomes a real-time mirror of the physical world.

[0046] S2: task triggering. When an intelligent conveying bottle completes its load conveying task at a certain work site, the medical staff takes it out from the receiving port. When the medical staff puts the empty bottle into the sending port of the site to prepare for recycling, the sensor combination (for example, the infrared reflection sensor detects that an object is put in, and at the same time the mechanical microswitch is pressed) in the site will capture this "put-in" behavior. The spatiotemporal correlation is performed between the physical event signal and the bottle ID read by the wireless network reader-writer at the entrance and exit of the site within the time window. After the data is uploaded to the intelligent scheduling system, the system can accurately determine that "bottle P001 is in the state of 'empty load to be recycled' at Site A at T1 time". This determination result formally creates an independent recycling task in the system, and triggers the next intelligent scheduling.

[0047] S3: Scheduling decision. This is the brain of the whole method, whose detailed decision flow is shown in Figure 5

[0048] S31: Candidate path generation. For the triggered recycling task (associated with transmission bottle i), the algorithm first acquires its current station (source point S) and the list of all recycling stations with current status "available" {D1, D2,..., Dm} (destination points) from the digital twin model. Then, one or more graph theory path planning algorithms are called. In a preferred embodiment, the classic **[Dijkstra algorithm]** or its improved version (such as A* algorithm if heuristic function can be estimated) is called to calculate the shortest path from source point S to each available destination point Dj, with physical distance or standard transmission time as the weight. To increase the richness of the decision, K shortest paths (K > 1) can also be calculated, forming a candidate path set {p1, p2,..., pK}.

[0049] S32: Multi-dimensional feature real-time quantification: The algorithm next quantifies the multi-dimensional features of each path pK in the candidate path set based on real-time data in the digital twin model.

[0050] (1) Path cost C(pK): This value can be directly the path length calculated by the path planning algorithm (unit: meters), or the theoretical transmission time estimated according to the standard speed. It represents the optimization goal of the shortest transmission time. To facilitate subsequent calculation, it is usually normalized, for example C_norm = C(pK) / C_max, where C_max is the maximum possible path length in the network.

[0051] (2) Task urgency U(i): This value reflects the task priority and deadline pressure of the recycling task. Its calculation can be a composite function, for example U(i) = w_p * P_i + w_t * f(T_now - T_start_wait). Where P_i is the static priority weight (between 0 and 1) queried in the configuration table according to the original task department of the bottle (such as bottles in emergency department or ICU have higher static priority) or the type of items once loaded in the bottle (such as blood samples have higher priority than ordinary medicines). T_start_wait is the timestamp when the bottle enters the "recycling waiting" state, and T_now is the current timestamp, the difference between the two represents the waiting time. The function f() can be a linear or nonlinear function (such as Sigmoid function) to represent the nonlinear relationship that the longer the waiting time, the faster the urgency grows. w_p and w_t are weight coefficients to balance the static priority and dynamic waiting time.

[0052] ​(3) Network congestion level L(pK): Path pK is composed of multiple pipe segments {seg1, seg2,...}. The real-time congestion level L_segment_j of each pipe segment seg_j can be provided by the digital twin model, defined as the number of transmission bottles currently carried by the pipe divided by its designed maximum capacity, or the flow rate per unit time. The comprehensive congestion level L(pK) of the path can be the average or maximum value of all pipe segments L_segment_j on the path, which quantifies the goal of avoiding waiting time and congestion phenomena.

[0053] (4) Resource energy consumption E(pK): Based on the total length of path pK, the total height of vertical lifting (work against gravity), the number and power of fans that need to be started, and the number of diverters that need to be switched, the system can estimate the theoretical energy consumption E(pK) for completing this transmission based on a preset physical energy consumption model. This is directly related to the resource constraints and green operation goals of the system.

[0054] S33: Weighted scoring and optimal decision. The system calculates the comprehensive score Score(pK) of each candidate path according to a dynamically configurable weighted scoring function (i.e. objective function F): (1) Dynamic weight coefficients (α, β, γ, δ): The sum of these coefficients is usually 1, representing the relative importance of different optimization goals, and can be dynamically adjusted. This gives the invention great flexibility. For example: 1) Time-of-day strategy: During the daytime business peak period (such as 9-11 am), the system can automatically increase the weight of γ (congestion avoidance); during the night business low period, it can increase the weight of δ (energy saving).

[0055] 2) Adaptive strategy based on system load: When the total number of "to be recycled" bottles in the system exceeds a certain threshold, the system can automatically increase the weight of β (task urgency) to prioritize processing bottles with the longest waiting time to alleviate overall system pressure.

[0056] (2) Finally, the algorithm selects the path pK* with the highest score as the optimal transmission path for this recycling task.

[0057] S34: Abnormality handling and decision rollback. During the decision-making process, the algorithm is built-in with rich abnormality handling logic. For example, if all available recycling stations are temporarily unreachable (as shown in Embodiment Three, cascading failures), or the comprehensive score of all candidate paths calculated is lower than a preset "execution threshold" (meaning the quality of all paths is very poor), the algorithm will not blindly issue instructions, but can put the transmission bottle in the "suspended scheduling" state, and wait for a preset time (such as 30 seconds) at its current location, and then trigger the calculation process again. This decision rollback mechanism avoids invalid or may cause more serious problems scheduling.

[0058] S4: Path execution. The calculated optimal transmission path pK* (which can be encoded as a set of instructions containing a series of time-synchronized diverter action sequences) is issued to the network control system, which is responsible for accurately driving the diverter to switch to the correct direction and starting the fan of the related pipe section to provide transmission power before the transmission bottle reaches the corresponding intersection with high real-time and high reliability.

[0059] S5: Automatic sorting. After the transmission bottle successfully reaches the designated recycling station along the planned path, it is received by the entrance guide section of the automatic sorting device. The device finally confirms the bottle identity through the RFID reader / writer at the entrance, and can optionally perform secondary inspection through the visual detection unit. Subsequently, the sorting executor sends the bottle into the correct recycling bin according to the instructions.

[0060] S6: Closed-loop confirmation. After the sensor (such as an infrared photoelectric sensor or a weighing sensor) in the recycling bin detects that a new bottle has entered, it sends a confirmation signal to the local controller of the automatic sorting device, which is uploaded to the central intelligent scheduling system. After receiving this final confirmation signal, the system updates the database state of the transmission bottle to "recycled - waiting for cleaning", and releases all temporary resources occupied by it in the scheduling model. At the same time, the full-process data of this recycling task (including path, time consumption, timestamps at each stage, etc.) will be archived to the historical database for subsequent performance analysis and algorithm optimization. Thus, a complete, intelligent, and automated recycling process realizes double closed-loop of data and physical layer.

Claims

1. A gas cylinder flow transfer bottle auto-return system, characterized by, The application relates to a smart logistics system for recycling waste bottles, comprising the following components: a plurality of smart transport bottles, each of which is integrated with an Internet of Things chip for storing unique identification information of the smart transport bottle; a sensing system composed of a plurality of wireless network read-write devices deployed in a gas logistics transmission network, which is used for automatic identification and positioning and real-time reading of information of the Internet of Things chip of the smart transport bottle passing by; a central intelligent scheduling system in communication with the sensing system, which is used for receiving and processing data uploaded by the wireless network read-write devices and internally storing an intelligent scheduling algorithm capable of dynamically planning an optimal transmission path for the smart transport bottle in an empty state according to real-time data; a network control system connected with the central intelligent scheduling system, which is used for receiving the optimal transmission path instruction and driving an execution mechanism in the gas logistics transmission network to guide the smart transport bottle to travel along the path; at least one automatic sorting device deployed at a designated recycling site, which is used for automatically physically classifying and arranging the arriving smart transport bottle according to the instruction of the central intelligent scheduling system or preset rules and placing the smart transport bottle in a corresponding recycling box.

2. The automatic return-to-stock system for gas cylinder flow transfer bottles according to claim 1, characterized by, The intelligent scheduling algorithm running in the central intelligent scheduling system is a heuristic algorithm based on multi-target weighted scoring.

3. The automatic return-to-stock system for gas cylinder flow transfer bottles according to claim 2, characterized by, The heuristic algorithm at least quantifies and comprehensively evaluates the following dimensions of characteristics: path cost, task priority, network congestion degree and energy consumption under resource constraints when making decisions.

4. The automatic return-to-stock system for gas cylinder flow transfer bottles according to claim 3, characterized by The task priority is dynamically calculated according to static task attributes associated with the smart transport bottle and waiting time after the smart transport bottle enters a recycling state.

5. The automatic gas cylinder transfer return system of claim 3, wherein, The central intelligent scheduling system can dynamically adjust the weight coefficient for calculating the multi-target weighted score according to a preset strategy to cope with congestion phenomena.

6. The automatic return-to-stock system for gas cylinder flow transfer bottles according to claim 1, characterized by, The Internet of Things chip is a passive ultra-high frequency radio frequency identification anti-metal flexible label, which is used for dynamically recording the carrying state and clean state of the smart transport bottle.

7. The automatic gas cylinder transfer return system of claim 1, wherein, The automatic sorting device comprises a wireless radio frequency identification read-write device, a main conveying belt and a sorting executor driven by an automatic control system, the wireless radio frequency identification read-write device is installed at the inlet of the main conveying belt and is used for reading information of the Internet of Things chip on the smart transport bottle.

8. The automatic return-to-stock system for gas cylinder flow transfer bottles according to claim 7, characterized by, The automatic sorting device is also integrated with a visual detection unit of high-efficiency identification technology, which is used for detecting the physical integrity of the smart transport bottle before sorting.

9. A gas cylinder flow transfer bottle automatic return method, characterized by, The application further discloses a recycling method of the smart logistics system, comprising the following steps: S1: state sensing, real-time monitoring of position information and state information of each smart transport bottle through the sensing system; S2: task triggering, automatically triggering a recycling task when it is detected that the smart transport bottle completes a transmission task and enters an empty state for recycling; S3: scheduling decision, calculating an optimal transmission path for the recycling task by the central intelligent scheduling system through the internally-stored intelligent scheduling algorithm; S4: path execution, the central intelligent scheduling system sends the optimal transmission path instruction to the network control system, which drives an execution mechanism in the gas logistics transmission network to guide the smart transport bottle to travel along the path; S5: automatic sorting, automatically physically classifying and arranging the smart transport bottle by the automatic sorting device when the smart transport bottle arrives at a designated recycling site. S6: Closed-loop confirmation, after the smart transport bottle is sorted, the central intelligent scheduling system confirms that the recycling task is completed, and updates its state information.

10. The method of claim 9, wherein, In the scheduling decision step, the intelligent scheduling algorithm makes decisions by the following ways: Calculate multiple candidate paths through path planning algorithm; Quantitative evaluation of multi-dimensional characteristics of each candidate path, including at least path cost, task priority, network congestion degree and energy consumption under resource constraints; Calculate and compare the comprehensive scores of each candidate path through multi-objective weighted scoring function, and select the path with the highest score as the optimal transport path.

Images