Nodal graph and reinforcement-learning model based systems and methods for managing moving agents
A method for managing moving agents is provided. The method comprises identifying goods agents, moving agents, and a plurality of requests within a predetermined area, generating a nodal graph including the moving agents, requests, and goods as vertices and edges defining relations between two vertices, obtaining actions for the moving agents by inputting the nodal graph to a reinforcement-learning based graphical neural network model stored in the moving agents, the reinforcement-learning based graphical neural network outputs the action for the moving agent in response to receiving the nodal graph, and instructing the moving agents to operate based on the actions to satisfy at least one of the plurality of requests.
1 . A method for managing moving agents, comprising:
identifying, by a controller of an autonomous vehicle, goods agents, moving agents, and a plurality of requests within a predetermined area;
generating, by the controller of the autonomous vehicle, a nodal graph including a plurality of vertices and a plurality of edges, the plurality of vertices including one or more vertices for the moving agents, one or more vertices for the requests, and one or more vertices for the goods agents, and each of the plurality of edges defining a relation between two vertices;
obtaining, by the controller of the autonomous vehicle, actions for the moving agents by inputting the nodal graph to a reinforcement learning based graphical neural network model stored in one of the moving agents, the reinforcement learning based graphical neural network model outputs the actions for the moving agents in response to receiving the nodal graph; and
instructing, by the controller of the autonomous vehicle, the moving agents to autonomously drive based on the actions to satisfy at least one of the plurality of requests.
2 . The method of claim 1 , wherein the moving agents are autonomous vehicle or human operated intelligent vehicles.
3 . The method of claim 1 , wherein the reinforcement learning based graphical neural network model outputs the actions that maximize a number of requests to be satisfied by the moving agents.
4 . The method of claim 1 , wherein the edges include a first edge between a vertex of a moving agent and an additional vertex of a request.
5 . The method of claim 1 , wherein the edges include a second edge between a vertex of a request and an additional vertex of an additional request, the second edge indicating that a moving agent can fulfill both the request and the additional request.
6 . The method of claim 1 , wherein the edges include at least one third edge between a vertex of a moving agent and an additional vertex of a goods agent.
7 . The method of claim 1 , wherein the reinforcement learning based graphical neural network model outputs the actions that minimize a sum of freight delays, scheduled requests delays, and preference costs.
8 . The method of claim 1 , further comprising trimming the nodal graph based on constraints on goods of the goods agents and services of the requests.
9 . The method of claim 8 , further comprising inputting the trimmed nodal graph to the reinforcement learning based graphical neural network model.
10 . The method of claim 4 , wherein the first edge indicating that the moving agent is able to fulfill the request.
11 . A system for managing moving agents, comprising:
controllers associated with the moving agents, each of the controllers configured to:
identify goods agents, the moving agents, and a plurality of requests within a predetermined area;
generate a nodal graph including a plurality of vertices and a plurality of edges, the plurality of vertices including one or more vertices for the moving agents, one or more vertices for the requests, and one or more vertices for the goods agents and each of the plurality of edges a defining relation between two vertices;
obtain actions for the moving agents by inputting the nodal graph to a reinforcement-learning based graphical neural network model stored in one of the moving agents, the reinforcement learning based graphical neural network model outputs the actions for the moving agents in response to receiving the nodal graph; and
instruct the moving agents to autonomously drive based on the actions to satisfy at least one of the plurality of requests.
12 . The system of claim 11 , wherein the moving agents are autonomous vehicle or human operated intelligent vehicles.
13 . The system of claim 11 , wherein the reinforcement learning based graphical neural network model outputs the actions that maximize a number of requests to be satisfied by the moving agents.
14 . The system of claim 11 , wherein the edges include a first edge between a vertex of a moving agent and an additional vertex of a request.
15 . The system of claim 11 , wherein the edges include a second edge between a vertex of a request and an additional vertex of an additional request, the second edge indicating that a moving agent can fulfill both the request and the additional request.
16 . The system of claim 11 , wherein the edges include at least one third edge between a vertex of a moving agent and an additional vertex of a goods agent.
17 . The system of claim 11 , wherein the reinforcement learning based graphical neural network model outputs the actions that minimize a sum of freight delays, scheduled requests delays, and preference costs.
18 . The system of claim 11 , wherein the controllers are further configured to trim the nodal graph based on constraints on goods of the goods agents and services of the requests.
19 . The system of claim 18 , wherein the controllers are further configured to input the trimmed nodal graph to reinforcement learning based graphical neural network model.
20 . The system of claim 14 , wherein the first edge indicating that the moving agent is able to fulfill the request.