CN-121995967-A - Multi-machine collaborative operation anti-collision monitoring system and method

CN121995967ACN 121995967 ACN121995967 ACN 121995967ACN-121995967-A

Abstract

The invention belongs to the technical field of intelligent control, and discloses a multi-machine collaborative operation anti-collision monitoring system and method, wherein the system comprises a distributed sensing layer, a local situation sensing layer and a local situation sensing layer, wherein the distributed sensing layer is used for acquiring local sensing data of current equipment and neighborhood sensing data of other equipment in a preset communication range, and a local situation sensing result is formed after space-time alignment and data integration; the system comprises a local situation sensing result, a cooperative cognition layer, a collision prediction layer and a control layer, wherein the local situation sensing result is used for executing data registration, eliminating coordinate deviation and time error among devices, constructing a global environment map, and building a group motion intention model according to the operation task type, the motion mode and the historical track of each device.

Inventors

FU HAIYANG
YANG DONGXU
ZHANG SHUAI

Assignees

济宁矿业集团有限公司霄云煤矿

Dates

Publication Date: 20260508
Application Date: 20251223

Claims (10)

1. The utility model provides a multi-machine collaborative operation anticollision monitored control system which characterized in that includes: the distributed sensing layer is used for acquiring local sensing data of the current equipment and neighborhood sensing data of other equipment in a preset communication range, and forming a local situation sensing result after space-time alignment and data integration; The collaborative cognitive layer performs data registration on the local situation sensing result, eliminates coordinate deviation and time error among devices, constructs a global environment map, and establishes a group movement intention model according to the operation task type, the movement mode and the history track of each device; The collision prediction layer predicts a device motion track cluster in a future time window based on the current motion state and the group motion intention model of each device, identifies track intersection, operation area overlapping and speed conflict conditions through space-time occupation modeling, calculates the collision probability among the devices and the time difference reaching a conflict point, and constructs a global collision risk matrix; The negotiation decision layer determines an avoidance master-slave relationship according to the global collision risk matrix, and generates a cooperative avoidance scheme; And the execution monitoring layer issues an avoidance command to the equipment according to the cooperative avoidance scheme, verifies the track of the equipment after the avoidance command is executed in real time, and triggers online rescheduling when the execution deviation is detected to drive the avoidance command to be effectively executed.
2. The system for monitoring and controlling collision of multiple machines in cooperation according to claim 1, wherein the method for obtaining the local perception data of the current device and the neighborhood perception data of other devices within the preset communication range comprises the following steps: The method comprises the steps that local perception data of the current equipment, including position coordinates, speed vectors, acceleration vectors, attitude angle information and obstacle point cloud information, are acquired in real time through a mounted positioning device, a motion state acquisition device and an environment perception sensor, are packaged into a perception data packet according to a unified format, and are additionally provided with an acquisition time stamp; The neighborhood aware data is provided by neighborhood devices within a preset communication range of the current device, consistent with the local aware data in data type.
3. The multi-machine collaborative operation anti-collision monitoring system according to claim 2, wherein the method for obtaining the local situation awareness result comprises: Converting position coordinates, attitude angle information and obstacle point cloud information contained in the neighborhood sensing data into a space coordinate system used by the current equipment, and synchronously carrying out attitude unification correction on a speed vector and an acceleration vector so as to realize space alignment of the neighborhood sensing data; Based on an acquisition time stamp attached to the local perception data, carrying out time correction on the neighborhood perception data after spatial alignment, and compensating sampling frequency differences among different devices through time interpolation to ensure that the reference time of the local perception data and the neighborhood perception data is kept uniform; and integrating the neighborhood sensing data after the space alignment and the time correction with the local sensing data to form a local situation sensing result.
4. A multi-machine collaborative operation anti-collision monitoring system according to claim 3, wherein the method for constructing a global environment map comprises: The method comprises the steps of extracting static reference features from local situation sensing results, establishing a cross-equipment unified reference feature set, identifying overlapping parts of observation areas among different devices according to the spatial coverage range of the local situation sensing results, performing feature matching on the reference feature set in the overlapping areas to obtain observation deviations, performing registration residual optimization based on the observation deviations, and performing multi-source superposition on the static reference features subjected to registration residual optimization to generate a global environment map.
5. The multi-machine collaborative operation anti-collision monitoring system according to claim 4, wherein the method for establishing a group exercise intent model comprises: collecting the operation task type, the motion mode and the history track information of each device, and carrying out feature extraction and normalization processing on the motion track and the motion mode of each device; According to the historical track, the motion mode and the job task type of the equipment, deducing possible motion targets and behavior modes of the equipment in a future time window to form motion intention labels of each equipment, and aggregating the motion intention labels of all the equipment to construct a group motion intention model.
6. The multi-machine collaborative operation anti-collision monitoring system according to claim 5, wherein the method for acquiring the equipment motion trail cluster comprises the following steps: respectively extracting current position coordinates, speed vectors, acceleration vectors and attitude angle information of each device from situation sensing results to form motion state vectors of the devices, and further obtaining the current motion state of each device; initializing a predicted time sequence by taking the current moment as a starting point, taking the current motion state vector of the equipment as an input initial value of a group motion intention model, and acquiring a motion intention label corresponding to the equipment; According to the current motion state and motion intention labels of each device, a prediction model is automatically selected from a constant speed model, a constant acceleration model and an interactive multi-model algorithm; based on the selected prediction model, the motion state vector at the current moment is taken as an initial state, the predicted position points at all moments in a future time window are obtained through gradual iterative calculation according to a prediction time sequence to form a predicted motion track, and the predicted motion tracks of all devices are collected to form a device motion track cluster.
7. The multi-machine collaborative operation anti-collision monitoring system according to claim 6, wherein the method for constructing a global collision risk matrix comprises: Based on a unified coordinate system of the equipment operation area, dividing the space area into grid units with fixed sizes, and sampling intervals in a time dimension according to the equipment movement track Mapping predicted position points contained in the predicted motion trail of each device to corresponding grids, and recording space occupation information of the device in different time slices in the future; and carrying out pair-by-pair intersection analysis on the predicted motion tracks of all the devices, identifying the conditions of track intersection, operation area overlapping and speed conflict, calculating the collision probability of each pair of devices, and constructing a global collision risk matrix.
8. The multi-machine collaborative operation anti-collision monitoring system according to claim 7, wherein the method for generating a collaborative avoidance scheme comprises: The collision probability in the global collision risk matrix is subjected to risk level sequencing, and a conflict device pair needing to be processed preferentially is determined; Determining a master-slave relationship of conflict equipment according to the constraint interval of the maneuverability and the path maintenance requirement, so as to determine master equipment and slave equipment, enabling the master equipment to maintain the original working path or task state, and enabling the slave equipment to execute avoidance operation; and generating a cooperative avoidance scheme for the slave device aiming at conflict types of track crossing, operation area overlapping and speed conflict.
9. The multi-machine collaborative operation collision avoidance monitoring system according to claim 8, wherein the method for efficiently executing the drive avoidance command comprises: According to the cooperative avoidance scheme, an avoidance command is issued to the equipment, wherein the avoidance command comprises a course adjustment amount, a speed change amount or a detour path point set, and the slave equipment executes a maneuvering action according to the avoidance command; The execution monitoring layer transmits deviation information back to the cooperative cognition layer, the collision prediction layer recalculates the motion track cluster and collision risk in the future time window, the negotiation decision layer generates a new cooperative avoidance scheme and dynamically updates the execution content of the avoidance instruction, and the avoidance instruction is driven to be effectively executed.
10. A multi-machine collaborative operation anti-collision monitoring method implemented by the multi-machine collaborative operation anti-collision monitoring system according to any one of claims 1 to 9, comprising: S1, acquiring local perception data of current equipment and neighborhood perception data of other equipment in a preset communication range, and forming a local situation perception result after space-time alignment and data integration; s2, performing data registration on the local situation sensing result, eliminating coordinate deviation and time error among cross-equipment, constructing a global environment map, and establishing a group movement intention model according to the operation task type, the movement mode and the historical track of each equipment; S3, predicting a device motion track cluster in a future time window based on the current motion state and the group motion intention model of each device, identifying track intersection, operation area overlapping and speed conflict conditions through space-time occupation modeling, calculating the collision probability among the devices and the time difference reaching a conflict point, and constructing a global collision risk matrix; s4, determining an avoidance master-slave relationship according to the global collision risk matrix, and generating a cooperative avoidance scheme; S5, issuing an avoidance command to the equipment according to the cooperative avoidance scheme, verifying the track of the equipment after executing the avoidance command in real time, triggering online rescheduling when the execution deviation is detected, and driving the avoidance command to be effectively executed.

Description

Multi-machine collaborative operation anti-collision monitoring system and method Technical Field The invention relates to the technical field of intelligent control, in particular to a multi-machine collaborative operation anti-collision monitoring system and method. Background With the increasing application of unmanned aerial vehicle cluster, mobile robot cluster and many autonomous devices collaborative operation, many devices simultaneously execute inspection, transportation, detection or construction tasks in shared space have become normal. However, due to the large number of devices and complex movement patterns, the existing multi-machine collaborative operation anti-collision technology still has the following problems: The global situation awareness capability is lacking, namely the existing system relies on a local sensor of a single device for environmental awareness, and a data sharing and neighborhood cooperative mechanism is lacking among devices, so that awareness information fragmentation and visual field limitation are caused, consistent global situation awareness cannot be formed, and the accuracy and reliability of overall cooperation are affected. The future track prediction capability is insufficient, namely the traditional prediction method cannot comprehensively consider behavior factors such as a job task, a historical track, a motion mode and the like, and lacks the combined deduction capability of the future states of multiple devices, so that the prediction deviation of potential conflict positions, conflict time and conflict probability is larger. The avoidance decision lacks global coordination, namely the prior art mostly adopts a local triggering type avoidance strategy, cannot integrally sort conflict risks of multiple devices, cannot reasonably distribute avoidance responsibility, is easy to avoid simultaneously, mutually interfere or secondarily conflict, and reduces the cooperation efficiency. In view of the above, the present invention provides a multi-machine collaborative anti-collision monitoring system and method for solving the above problems. Disclosure of Invention In order to overcome the defects in the prior art, the invention provides the following technical scheme that the multi-machine collaborative operation anti-collision monitoring system comprises: the distributed sensing layer is used for acquiring local sensing data of the current equipment and neighborhood sensing data of other equipment in a preset communication range, and forming a local situation sensing result after space-time alignment and data integration; The collaborative cognitive layer performs data registration on the local situation sensing result, eliminates coordinate deviation and time error among devices, constructs a global environment map, and establishes a group movement intention model according to the operation task type, the movement mode and the history track of each device; The collision prediction layer predicts a device motion track cluster in a future time window based on the current motion state and the group motion intention model of each device, identifies track intersection, operation area overlapping and speed conflict conditions through space-time occupation modeling, calculates the collision probability among the devices and the time difference reaching a conflict point, and constructs a global collision risk matrix; The negotiation decision layer determines an avoidance master-slave relationship according to the global collision risk matrix, and generates a cooperative avoidance scheme; And the execution monitoring layer issues an avoidance command to the equipment according to the cooperative avoidance scheme, verifies the track of the equipment after the avoidance command is executed in real time, and triggers online rescheduling when the execution deviation is detected to drive the avoidance command to be effectively executed. Preferably, the method for obtaining the local awareness data of the current device and the neighborhood awareness data of other devices within the preset communication range includes: The method comprises the steps that local perception data of the current equipment, including position coordinates, speed vectors, acceleration vectors, attitude angle information and obstacle point cloud information, are acquired in real time through a mounted positioning device, a motion state acquisition device and an environment perception sensor, are packaged into a perception data packet according to a unified format, and are additionally provided with an acquisition time stamp; The neighborhood aware data is provided by neighborhood devices within a preset communication range of the current device, consistent with the local aware data in data type. Preferably, the method for obtaining the local situation awareness result includes: Converting position coordinates, attitude angle information and obstacle point cloud information contained in the neighborho