A multi-agent scheduling control method and system for isolation wards

Abstract

The application provides a multi-agent scheduling control method and system for isolated wards, and the method comprises the following steps: establishing a multi-isolated area target task point request distribution scheduling target function based on multi-agents; each isolated area target task point submits a distribution request to the system according to the demand; and the system schedules the agents to complete the distribution task according to the scheduling target function. The group control scheduling method can solve the problem of unreasonable scheduling and distribution, realize the function of scheduling the multi-agent distribution task in combination with the distribution scheduling system, optimize the defects of the traditional scheduling multi-agent method, such as long time consumption and large energy consumption, and improve the distribution efficiency of the multi-agent in the multi-task and multi-area intensive environment.

Description

Technical Field

[0001] This invention relates to the field of multi-agent scheduling and control technology, and in particular to a multi-agent scheduling and control method and system for isolation wards. Background Technology

[0002] Currently, various methods have been applied to multi-agent scheduling tasks, including ant colony optimization, simulated annealing, genetic algorithms, and A* algorithms. All of these methods aim to find the optimal solution with the shortest path and the least time consumption under certain constraints.

[0003] CN 113739812 A discloses a delivery plan generation method, system, system, and computer-readable storage medium. The method includes the following steps: obtaining information related to multiple routes between multiple locations, information related to the types of multiple delivery vehicles, and information related to traffic conditions in multiple time periods corresponding to the types of delivery vehicles; obtaining multiple cost matrices including the travel time required to travel the multiple routes in each time period; calculating the travel time for delivering a first delivery task via a first route and a second route between a first delivery location and a second delivery location, based on the travel time in the first time period and the second time period; and comparing the travel time for delivering the first delivery task via each route to determine the route for delivering the first delivery task. However, this disclosed method only considers travel time costs and does not consider the acceleration energy consumption of delivery vehicles.

[0004] Publication number CN 110782064 A discloses a visualization method and system integrating vehicle scheduling optimization and task allocation. The method includes the following steps: receiving customer orders and collecting data information required to construct a visualization map; an algorithm server establishes a corresponding vehicle scheduling model based on customer needs, determines the vehicle scheduling model through the transportation costs required for cargo transportation, and establishes an objective function with the goal of minimizing the total cost; the algorithm server calculates the optimal vehicle scheduling scheme; based on the optimal scheduling scheme provided by the algorithm server, the route map of the transport vehicles is visualized on Baidu Maps; the upper-level scheduling system sends the tasks of each depot and each vehicle, as well as the specific driving routes of the vehicles, to the corresponding drivers at the lower level, and the drivers deliver goods according to the arrangements; using the GPS positioning function provided by Baidu Maps, the system performs real-time location tracking and location feedback for each vehicle; vehicle drivers report the task completion progress, vehicle status, congestion conditions on various road sections, and emergencies to the upper-level planning layer in a timely manner through mobile terminals, so that the upper-level planning layer can optimize and adjust the next task in a timely manner based on the actual situation; after the task is completed, the system automatically updates the status of each depot and each vehicle, as well as the usage of materials, and stores customer information and feedback on this delivery for the next scheduling arrangement. However, the publicly available method only considers the cost of vehicle travel distance, the time penalty cost for early or delayed delivery, and the fixed cost of using the vehicle, without considering the acceleration energy consumption of the delivery vehicle.

[0005] The currently disclosed methods have the following shortcomings: 1) They do not consider the energy consumption required for agent acceleration and deceleration, which leads to increased energy consumption of the agents; 2) When the agent approaches the target task point at a relatively high speed, the task signals of nearby target task points are usually shielded to avoid the scheduling model algorithm calculating infinite or infinitesimal results, which also leads to a waste of multi-agent resources and reduces scheduling efficiency. Although current multi-agent scheduling methods have been widely used in many fields, the currently disclosed multi-agent scheduling methods cannot be directly applied to delivery tasks in the dense environment of multiple tasks and multiple areas in isolation wards due to the above shortcomings. Summary of the Invention

[0006] To address the aforementioned issues, this invention provides a multi-agent scheduling and control method and system for isolation wards, thereby overcoming the shortcomings of existing technologies in multi-agent scheduling and control in isolation wards, which suffer from low efficiency and high energy consumption, and achieving more efficient scheduling of multi-agents.

[0007] To achieve this objective, the present invention adopts the following technical solution:

[0008] The first aspect of this application provides a multi-agent scheduling and control method for isolation wards, including the following steps:

[0009] S11. The distribution center system traverses each target task point and obtains the delivery task signals sent by the target task points.

[0010] S12. Determine if there is a delivery task signal at the target task point:

[0011] If there is no pending delivery task signal at the target task point, execute S11; if there is a pending delivery task signal at the target task point, execute S13.

[0012] S13. The distribution center system obtains the current location and speed information of all available intelligent agents;

[0013] S14. The distribution center system inputs the current position and speed information of all available agents into the multi-agent delivery scheduling objective function.

[0014] The objective function for agent delivery scheduling is:

[0015]

[0016] in:

[0017] C d The unit delivery distance coefficient for the intelligent agent;

[0018] C n The startup energy consumption coefficient of the intelligent agent;

[0019] C m Task factor coefficients have been assigned to the agent;

[0020] M k M represents the number of tasks assigned to the agent. k =1; 2; ...; a, where a is the maximum number of intelligent agent storage slots;

[0021] C v The deceleration energy consumption coefficient of the intelligent agent;

[0022] v k The current velocity of the agent;

[0023] The agent's current acceleration;

[0024] D i，k The distance between the agent and the target task point;

[0025] i is the target task point number, i = 1; 2; ...; n, where n is the number of target task points;

[0026] k is the agent number, k = 1; 2; ...; b, where b is the number of agents;

[0027] ω1 is the weighting coefficient of the distance factor;

[0028] ω2 is the startup energy consumption weighting coefficient;

[0029] ω3 is the weighting coefficient for the number of assigned tasks;

[0030] ω4 is the deceleration energy consumption weighting coefficient;

[0031] ω j j = 1; 2; 3; 4 are the preset weight coefficients in the multi-agent delivery scheduling objective function, ω j The constraints are shown in formula (2):

[0032]

[0033] ω1C d D i，k Let $\frac{k}{i}$ be the distance cost from the $k-th agent to the $i-th target task point.

[0034] The startup energy cost of the kth agent;

[0035] ω3C m M k Cost of the number of tasks assigned to the kth agent;

[0036] Let $\frac{k}{i}$ be the energy cost of deceleration from the $k-th agent to the $i-th target task point.

[0037] f(i, k) is the objective function of the k-th agent to the i-th target task point;

[0038] S15. Determine whether the distribution center system has traversed all available agents:

[0039] If the distribution center system has not traversed all available agents, then execute S14; if the distribution center system has traversed all available agents, then execute S16.

[0040] S16. The distribution center system compares the objective function values ​​of each available agent.

[0041] S17. Determine whether the distribution center system has compared the objective function values ​​of all available agents:

[0042] If the objective function values ​​of all available agents have not been completely compared, then execute S16; if the agent with the smallest objective function value among all available agents has been found, then execute S18.

[0043] S18. Assign the delivery task signal of the current target task point to the agent with the smallest calculated objective function value, and define it as the pre-tuned agent.

[0044] S19. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent:

[0045] If the distribution center system does not receive a confirmation signal from the pre-adjusted agent, then execute S20; if the distribution center system has received a confirmation signal from the pre-adjusted agent, then execute to end the current task scheduling and allocation process.

[0046] S20. The distribution center system sends delivery tasks to pre-selected agents through other agents;

[0047] S21. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent:

[0048] If the distribution center system still has not received the confirmation signal returned by the pre-adjusted agent, then execute S22; if the distribution center system has received the confirmation signal returned by the pre-adjusted agent, then execute to end the current task scheduling and allocation process.

[0049] S22. Record the current pre-adjusted agent as a faulty agent and then execute S13.

[0050] The second aspect of this application provides a multi-agent scheduling and distribution center system for isolation wards. The distribution center system includes: a wireless network module for communicating and configuring with agents, receiving status information of agents and target task points, and sending control agent information;

[0051] The model calculation module is used to input the state information of the agent and the target task point into the multi-agent delivery scheduling objective function and calculate the output function value;

[0052] The storage module is used to store fault information of the intelligent agent;

[0053] The agents that interact with the distribution center system include:

[0054] The wireless network module is used to communicate with the distribution center system and other intelligent agents, send the current location and speed information of the intelligent agents, and receive scheduling information from the distribution center system.

[0055] The environmental perception module is used to obtain the current position and speed information of the intelligent agent;

[0056] Power output module, used for mobile intelligent agents;

[0057] The target task points for interacting with the distribution center system include:

[0058] The wireless network module is used for communication configuration with the distribution center system and for sending delivery task signals to the distribution center system.

[0059] Compared with existing technologies, the present invention has the following advantages and effects:

[0060] (1) In the objective function of scheduling task allocation in the isolation ward, the present invention introduces the start-up energy consumption cost term and the deceleration energy consumption cost term, which can reduce the energy consumption of the intelligent agent.

[0061] (2) The scheduling task allocation objective function proposed in this invention can be applied to global all-time scheduling tasks. The deceleration energy consumption cost term in the objective function can improve the task allocation rate when the agent approaches the target task point, thereby improving the efficiency of global scheduling. Attached Figure Description

[0062] To more clearly explain the technical structure of the embodiments of the present invention, the following is a brief description of the drawings necessary for the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention, and other drawings can be derived from these drawings without creative effort.

[0063] Figure 1 This is a flowchart illustrating the multi-agent scheduling and control method for isolation wards provided by the present invention.

[0064] Figure 2 This is a diagram of the operating environment architecture of the multi-agent scheduling and control method for isolation wards provided by the present invention.

[0065] Figure 3 This is a functional module diagram of the distribution center system provided by the present invention.

[0066] Figure 4 This is a functional block diagram of the intelligent agent provided by the present invention.

[0067] Figure 5 This is a functional module diagram of the target task points provided by the present invention. Detailed Implementation

[0068] To enable those skilled in the art to better understand the present invention, the embodiments of the present invention will be clearly and completely described below with accompanying drawings according to one embodiment. Obviously, these described embodiments are only a part of the present invention and not all embodiments. According to the embodiments of the present invention, other implementations obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0069] This invention proposes a multi-agent scheduling and control method and system for isolation wards, which is described in detail below with reference to the accompanying drawings:

[0070] Figure 1 This is a flowchart illustrating the multi-agent scheduling and control method for isolation wards provided by the present invention. Figure 1 As shown, the multi-agent scheduling and control method for isolation wards specifically includes:

[0071] S11. The distribution center system traverses each target task point and obtains the delivery task signals sent by the target task points.

[0072] S12. Determine if there is a delivery task signal at the target task point:

[0073] If there is no pending delivery task signal at the target task point, execute S11; if there is a pending delivery task signal at the target task point, execute S13.

[0074] S13. The distribution center system obtains the current location and speed information of all available intelligent agents;

[0075] S14. The distribution center system inputs the current position and speed information of all available agents into the multi-agent delivery scheduling objective function.

[0076] The objective function for agent delivery scheduling is:

[0077]

[0078] in:

[0079] C d The unit delivery distance coefficient for the intelligent agent;

[0080] C n The startup energy consumption coefficient of the intelligent agent;

[0081] C m Task factor coefficients have been assigned to the agent;

[0082] M k M represents the number of tasks assigned to the agent. k =1; 2; ...; a, where a is the maximum number of intelligent agent storage slots;

[0083] C v The deceleration energy consumption coefficient of the intelligent agent;

[0084] v k The current velocity of the agent;

[0085] The agent's current acceleration;

[0086] D i，k The distance between the agent and the target task point;

[0087] i is the target task point number, i = 1; 2; ...; n, where n is the number of target task points;

[0088] k is the agent number, k = 1; 2; ...; b, where b is the number of agents;

[0089] ω1 is the weighting coefficient of the distance factor;

[0090] ω2 is the startup energy consumption weighting coefficient;

[0091] ω3 is the weighting coefficient for the number of assigned tasks;

[0092] ω4 is the deceleration energy consumption weighting coefficient;

[0093] ω j j = 1; 2; 3; 4 are the preset weight coefficients in the multi-agent delivery scheduling objective function, ω j The constraints are shown in formula (2):

[0094]

[0095] ω1C d D i，k Let $\frac{k}{i}$ be the distance cost from the $k-th agent to the $i-th target task point.

[0096] The startup energy cost of the kth agent;

[0097] ω3C m M k Cost of the number of tasks assigned to the kth agent;

[0098] Let $\frac{k}{i}$ be the energy cost of deceleration from the $k-th agent to the $i-th target task point.

[0099] f(i, k) is the objective function of the k-th agent to the i-th target task point;

[0100] S15. Determine whether the distribution center system has traversed all available agents:

[0101] If the distribution center system has not traversed all available agents, then execute S14; if the distribution center system has traversed all available agents, then execute S16.

[0102] S16. The distribution center system compares the objective function values ​​of each available agent.

[0103] S17. Determine whether the distribution center system has compared the objective function values ​​of all available agents:

[0104] If the objective function values ​​of all available agents have not been completely compared, then execute S16; if the agent with the smallest objective function value among all available agents has been found, then execute S18.

[0105] S18. Assign the delivery task signal of the current target task point to the agent with the smallest calculated objective function value, and define it as the pre-tuned agent.

[0106] S19. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent:

[0107] If the distribution center system does not receive a confirmation signal from the pre-adjusted agent, then execute S20; if the distribution center system has received a confirmation signal from the pre-adjusted agent, then execute to end the current task scheduling and allocation process.

[0108] S20. The distribution center system sends delivery tasks to pre-selected agents through other agents;

[0109] S21. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent:

[0110] If the distribution center system still has not received the confirmation signal returned by the pre-adjusted agent, then execute S22; if the distribution center system has received the confirmation signal returned by the pre-adjusted agent, then execute to end the current task scheduling and allocation process.

[0111] S22. Record the current pre-adjusted agent as a faulty agent and then execute S13.

[0112] Figure 2 This is a diagram illustrating the runtime environment architecture of the multi-agent scheduling and control method for isolation wards provided by this invention. The technical solutions covered in this embodiment are as follows:

[0113] b intelligent agents 100 and n target task points 120 communicate with the distribution center system 110. The two-dimensional coordinates of the n target task points 120 are different. When a target task point 120 has a delivery requirement, it sends a delivery task signal to the distribution center system 110. After calculation, the distribution center system 110 obtains the objective function of each intelligent agent and sends the current delivery task signal to the intelligent agent 100 with the smallest objective function value.

[0114] Figure 3 This is a functional block diagram of the distribution center system provided by the present invention. The scope of the technical solutions covered in this embodiment is as follows:

[0115] The distribution center system described in this example, specifically the distribution center system 110, includes: a wireless network module 111, a model calculation module 112, and a storage module 113.

[0116] The wireless network module 111 is used by the distribution center system 110 to receive the delivery task signal from the target task point, the confirmation signal sent by the pre-adjusted intelligent agent, the position and speed information of each intelligent agent, and send the delivery task signal to the intelligent agent.

[0117] The model calculation module 112 includes a multi-agent delivery scheduling model for calculating the objective function of each agent 100. In this example, the model calculation module 112 is a computer-executable program that, when executed by a computer, implements a multi-agent scheduling control method for isolation wards. The model calculation module 112 takes the delivery task signal, agent position, and speed information received from the wireless network module 111 as input to the multi-agent delivery scheduling objective function. After calculation by the model calculation module 112, the delivery task signal is sent to the agents through the wireless network module 111.

[0118] Storage module 113 is used to store agent fault information recorded in the distribution center system. In this example, the storage module is a non-transitory storage medium. Model calculation module 112 directly sends delivery task signals to the pre-tuned agent through wireless network module 111, with a preset time threshold. If no confirmation signal is received from the agent, model calculation module 112 again sends delivery task signals to the pre-tuned agent through wireless network module 111 and other agents, with a preset time threshold. If no confirmation signal is received from the pre-tuned agent after the preset time, model calculation module 112 records the pre-tuned agent fault and stores the fault information in storage module 113. The time threshold can be set according to actual conditions.

[0119] Figure 4 This is a functional block diagram of the intelligent agent provided by the present invention. The intelligent agent 100 can be a mobile robot or a flying robot with sensing capabilities. The scope of the technical solutions covered in this embodiment is as follows:

[0120] The wireless network module 101 is used to receive delivery task signals from the distribution center system and send current location, speed information, and confirmation signals to the distribution center system. Preferably, the wireless network module 101 can transmit and receive signals via a flexible FPC antenna.

[0121] The environment perception module 102 is used to acquire distance information, speed information, and two-dimensional coordinate information of the surrounding three-dimensional scene of the intelligent agent 100. In this example, the environment perception module 102 includes a binocular camera and a lidar for acquiring distance information of the surrounding three-dimensional scene of the intelligent agent 100; the environment perception module 102 acquires the two-dimensional coordinate information of the intelligent agent 100 by performing indoor two-dimensional coordinate positioning through the wireless network module 101; and acquires the speed information of the intelligent agent 100 through the power output module 103.

[0122] The power output module 103 is used to move the intelligent agent 100 via a motor and obtain the motor speed and the speed of the intelligent agent 100. In this example, the intelligent agent 100 can be moved in the following ways: a tracked intelligent agent 100 driven by a motor, a quadcopter intelligent agent 100 driven by a motor, or an intelligent agent 100 with wheels driven by a motor; the power output module 103 obtains the motor speed through a coaxial encoder of the motor and obtains the speed of the intelligent agent 100 through an acceleration module installed on the main body of the intelligent agent 100.

[0123] Figure 5 This is a functional block diagram of the target task point provided by the present invention. The target task point 120 can be a ward, laboratory, or pharmacy. The scope of the technical solution covered in this embodiment is as follows:

[0124] Wireless network module 121 is used to send delivery task signals to target task point 120.

[0125] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A multi-agent scheduling and control method for isolation wards, characterized in that, Using the agent's speed, acceleration, current number of tasks, and distance to the target task point as variables, a multi-agent delivery scheduling objective function can be used to obtain the agent's real-time objective function value. This real-time objective function value serves as the task allocation indicator for multiple agents in the isolation ward. The method includes the following steps: S11. The distribution center system traverses each target task point and obtains the delivery task signals sent by the target task points. S12. Determine if there is a delivery task signal at the target task point: If there is no pending delivery task signal at the target task point, execute S11; if there is a pending delivery task signal at the target task point, execute S13. S13. The distribution center system obtains the current location and speed information of all available intelligent agents; S14. The distribution center system inputs the current position and speed information of all available agents into the multi-agent delivery scheduling objective function. The multi-agent delivery scheduling objective function is: (1) in: The unit delivery distance coefficient for the intelligent agent; The startup energy consumption coefficient of the intelligent agent; Task factor coefficients have been assigned to the agent; The number of tasks assigned to the agent. , This represents the maximum number of intelligent agent storage tanks. The deceleration energy consumption coefficient of the intelligent agent; The current velocity of the agent; The agent's current acceleration; The distance between the agent and the target task point; Number the target task points. , The number of target task points; Number the intelligent agent. , The number of agents; This refers to the distance factor weighting coefficient. To activate the energy consumption weighting factor; Weighting coefficient for the number of assigned tasks; This is the energy consumption weighting coefficient for deceleration; These are the preset weighting coefficients in the objective function for multi-agent delivery scheduling. The constraints are shown in formula (2): (2) For the first The agent to the first Distance cost of each target task point; For the first The startup energy cost of an individual agent; For the first The cost of the number of tasks assigned to each agent; For the first The agent to the first Energy consumption cost of deceleration at each target task point; For the first The agent to the first The objective function for each target task point; S15. Determine whether the distribution center system has traversed all available agents: If the distribution center system has not traversed all available agents, then execute S14; if the distribution center system has traversed all available agents, then execute S16. S16. The distribution center system compares the objective function values ​​of each available agent. S17. Determine whether the distribution center system has compared the objective function values ​​of all available agents: If the objective function values ​​of all available agents have not been completely compared, then execute S16; if the agent with the smallest objective function value among all available agents has been found, then execute S18. S18. Assign the delivery task signal of the current target task point to the agent with the smallest calculated objective function value, and define it as the pre-tuned agent. S19. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent: If the distribution center system does not receive a confirmation signal from the pre-adjusted agent, then execute S20; if the distribution center system has received a confirmation signal from the pre-adjusted agent, then execute to end the current task scheduling and allocation process. S20. The distribution center system sends delivery tasks to pre-selected agents through other agents; S21. Determine whether the distribution center system has received the confirmation signal returned by the pre-adjusted intelligent agent: If the distribution center system still has not received the confirmation signal returned by the pre-adjusted agent, then execute S22; if the distribution center system has received the confirmation signal returned by the pre-adjusted agent, then execute to end the current task scheduling and allocation process. S22. Record the current pre-adjusted agent as a faulty agent and then execute S13.

2. The method according to claim 1, characterized in that, The distribution center system includes: The wireless network module is used to communicate and configure with the intelligent agent, receive status information of the intelligent agent and the target task point, and send control information to the intelligent agent. The model calculation module is used to input the state information of the agent and the target task point into the multi-agent delivery scheduling objective function and calculate the output function value; The storage module is used to store fault information of the intelligent agent; The agents that interact with the distribution center system include: The wireless network module is used to communicate with the distribution center system and other intelligent agents, send the current location and speed information of the intelligent agents, and receive scheduling information from the distribution center system. The environmental perception module is used to obtain the current position and speed information of the intelligent agent; Power output module, used for mobile intelligent agents; The target task points for interacting with the distribution center system include: The wireless network module is used for communication configuration with the distribution center system and for sending delivery task signals to the distribution center system.

Images