CN-122018325-A - Sand sucking, filling and multi-arm cooperative control algorithm based on kinematic coupling

CN122018325ACN 122018325 ACN122018325 ACN 122018325ACN-122018325-A

Abstract

The invention provides a sand sucking and filling multi-arm cooperative control algorithm based on kinematic and dynamic coupling, which comprises the following steps of S1, construction of a full-working-condition data acquisition and feature library of a hydrodynamic interference field, S2, unified mapping of multi-arm kinematic modeling and working space, S3, construction of a multi-arm kinematic-dynamic coupling model embedded with a hydrodynamic interference self-adaptive compensation item, S4, multi-arm task-dynamic coupling distribution based on an improved parasitic-predation algorithm, S5, track-task bidirectional coupling planning based on a fractional order pseudo-spectrum method, S6, multi-arm-filling cooperative closed-loop control based on port controlled Hamiltonian fractional order passive control, and S7, sand sucking and filling parameters real-time acquisition and full-system closed-loop feedback. According to the invention, through the bidirectional collaborative optimization of the deep coupling modeling of kinematics and dynamics and the core control algorithm, the high-precision, high-stability and high-efficiency collaborative control of the multi-arm sand sucking and filling operation in the dynamic water area environment is realized.

Inventors

WANG LIHUA
LI BOFENG
GU YONG
MA ZHONGXIAN
ZHOU PENG

Assignees

中交上海航道局有限公司
同济大学

Dates

Publication Date: 20260512
Application Date: 20260408

Claims (10)

1. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic coupling is characterized by comprising the following steps of: S1, acquiring all-condition hydrodynamic interference data of a sand sucking and filling operation water area, and constructing a hydrodynamic interference field feature library after preprocessing the acquired all-condition hydrodynamic interference data; S2, constructing a single mechanical arm kinematics model for each mechanical arm in the multi-arm collaborative operation system, and simultaneously constructing a homogeneous transformation mapping relation between a base coordinate system and a global operation coordinate system of a plurality of mechanical arms to finish unified mapping of a multi-arm collaborative operation space; s3, constructing a multi-arm kinematics-dynamics coupling model embedded with a hydrodynamic disturbance compensation term based on a single mechanical arm kinematics model, a homogeneous transformation mapping relation and a hydrodynamic disturbance field feature library; s4, completing task-dynamic coupling distribution of sand sucking and filling operation of the multi-arm by adopting an improved parasitic-predation algorithm by taking the multi-arm kinematics-dynamic coupling model as hard constraint, so as to obtain an optimal task distribution scheme of the multi-arm; S5, completing track-task bidirectional coupling planning by adopting a fractional order pseudo-spectrum method based on a multi-arm optimal task allocation scheme and a multi-arm kinematic-dynamic coupling model, generating a multi-arm operation optimal track, and simultaneously feeding back a track planning result to the S4 optimal task allocation scheme; And S6, based on the optimal track of multi-arm operation and a multi-arm kinematics-dynamics coupling model, adopting a port-controlled Hamiltonian fractional order passive control algorithm to complete cooperative closed-loop control of multi-arm movement and sand suction and filling processes, and outputting a mechanical arm joint driving control instruction.
2. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic and dynamic coupling according to claim 1 is characterized by further comprising S7, synchronously collecting real-time data of sand sucking and filling operation through a multi-sensor fusion framework, performing filtering treatment on the collected real-time data, and feeding the processed real-time data back to a multi-arm kinematic-dynamic coupling model of S3, an improved parasitic-predation algorithm of S4, a fractional order pseudo-spectrum method of S5 and a port controlled Hamiltonian fractional order passive control algorithm of S6 through a full-link, so as to complete closed-loop dynamic optimization of the whole system.
3. The sand sucking and filling multi-arm cooperative control algorithm based on the coupling of kinematics and dynamics according to claim 1, wherein in the step S2, a single mechanical arm kinematics model is built by adopting a D-H parameter method, the construction of the homogeneous transformation mapping relation is completed based on the installation layout of a multi-arm cooperative operation system, and meanwhile, the effective operation range of the single mechanical arm kinematics model is limited by combining the coverage range of a hydrodynamic interference field feature library.
4. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic and dynamic coupling according to claim 1, wherein in the step S3, a multi-arm kinematic-dynamic coupling model is constructed by adopting a Newton-Euler recursion method, the mechanical arm joint motion parameters output by a single mechanical arm kinematic model are used as endogenous input variables of a dynamic equation, an embedded hydrodynamic disturbance compensation term is matched with the current working condition based on a hydrodynamic disturbance field feature library to complete real-time updating, and meanwhile, the kinematic coupling constraint and the forced coupling constraint among a plurality of mechanical arms are added into the multi-arm kinematic-dynamic coupling model.
5. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic and dynamic coupling according to claim 1, wherein in S4, the parasitic-predation algorithm is improved to minimize the comprehensive operation cost of the multi-arm system as an optimization target, the comprehensive operation cost covers the total operation execution time, the total motion path length, the total driving energy consumption and the total hydrodynamic interference compensation error, and the hard constraint conditions set by the parasitic-predation algorithm comprise joint moment constraint, joint motion constraint, hydrodynamic compensation constraint and collision safety constraint, and all constraint conditions are derived from a multi-arm kinematic-dynamic coupling model.
6. The sand sucking and filling multi-arm cooperative control algorithm based on the coupling of kinematics and dynamics according to claim 1, wherein in S4, the iterative process of improving the parasitic-predation algorithm sequentially executes individual position updating of predation phase, parasitic phase and symbiotic phase, wherein the symbiotic phase synchronously accesses the track planning result output by fractional pseudo-spectroscopy, adjusts the individual position updating strategy based on the track optimizing effect, and synchronously realizes bidirectional cooperative optimization of task allocation and track planning.
7. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic and dynamic coupling according to claim 1, wherein in the step S5, a fractional order pseudo-spectrum method is constructed based on a Caputo fractional order differential operator, the trajectory control energy consumption, the fractional order concussion amount and the task allocation deviation are minimized as optimization targets, and constraint conditions of trajectory optimization comprise fractional order dynamic constraints obtained through conversion based on a multi-arm kinematic-dynamic coupling model.
8. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic and dynamic coupling according to claim 1, wherein in S6, the port-controlled hamilton fractional order passive control algorithm is based on a hamilton energy function design control law of a mechanical arm system, the control law comprises a feedforward control item, a fractional order feedback control item and a robust compensation item, the feedforward control item is derived from an optimal track of multi-arm operation, and the robust compensation item is dynamically adjusted based on a hydrodynamic interference field feature library and load disturbance data acquired in real time.
9. The sand sucking and filling multi-arm cooperative control algorithm based on kinematic coupling according to claim 1 is characterized in that in S6, based on a reference state of real-time data and a multi-arm operation optimal track, a real-time tracking error maximum value in a whole operation period is calculated, the real-time tracking error maximum value is fed back to a fractional order pseudo-spectrum method of S5, smoothness constraint of track optimization is dynamically adjusted, real-time operation comprehensive cost data is calculated according to an objective function of an improved parasitic-predation algorithm based on real-time data, the real-time operation comprehensive cost data is fed back to S4, and on-line dynamic optimization of a task allocation scheme is triggered, so that two-way cooperative optimization of the improved parasitic-predation algorithm, the fractional order pseudo-spectrum method and a port controlled Hamiltonian fractional order passive control algorithm is achieved.
10. The sand sucking and filling multi-arm cooperative control algorithm based on the kinematic coupling according to claim 2, wherein in S7, the collected real-time data comprise sand sucking and filling process parameters, mechanical arm motion state parameters and working water area hydrologic parameters, the real-time load variation is calculated based on the collected sand sucking and filling process parameters, disturbance items of the multi-arm kinematic-kinematic coupling model constructed in S3 are corrected, and all the data after filtering processing are fed back to corresponding algorithm modules in real time, so that the online dynamic optimization and the full-working-condition self-adaption of the whole system are realized.

Description

Sand sucking, filling and multi-arm cooperative control algorithm based on kinematic coupling Technical Field The invention relates to the technical field of cooperative control of underwater engineering equipment and multiple robots, in particular to a sand sucking, filling and multi-arm cooperative control algorithm based on kinematic coupling. Background Along with the large-scale development of underwater engineering operations such as inland waterway dredging, offshore surrounding and filling engineering, water conservancy facility repair and the like, the working condition complexity and the operation precision requirements of sand sucking and filling operation are continuously improved, and the multi-arm collaborative operation system has become core equipment of sand sucking and filling operation by virtue of the advantages of wide operation range, strong loading capacity and flexible action. The existing multi-arm cooperative control technology forms a mature application scheme in a conventional industrial scene, but has obvious technical defects aiming at the sand sucking and filling operation scene of an underwater dynamic water area: Firstly, in the prior art, a modeling and control mode of kinematics and dynamics decoupling is mostly adopted, track planning is finished based on a kinematics model, then compensation and correction are carried out through a dynamics model, the problem that the planned track exceeds the dynamic feasible range of a mechanical arm is very easy to occur, and in the dynamic environment of water flow disturbance, joint overload and track tracking precision are further greatly reduced; Secondly, multi-arm task allocation mainly adopts traditional particle swarm optimization, genetic algorithm and other swarm intelligent algorithms, only uses operation time and path length as optimization targets, and does not embed mechanical arm dynamics constraint and hydrodynamic interference influence into the optimization process, so that the obtained allocation scheme has the problems of high energy consumption, interference compensation failure, multi-arm cooperative conflict and the like in actual operation; Thirdly, a serial processing mode is adopted for track planning and task allocation, the track planning can be completed only based on a given task allocation result, and a task allocation scheme cannot be reversely optimized through the feasibility of the track planning, so that the task allocation and the track planning are disjointed, and the overall operation performance cannot reach global optimum; Fourthly, the closed-loop control of the underwater operation mostly adopts algorithms such as conventional PID, sliding mode control and the like, the robustness and the anti-interference capability are insufficient under the working conditions of strong disturbance of dynamic water flow and real-time change of sand suction load, the motion control of the mechanical arm and the concentration and flow control of sand suction filling are mutually independent, the cooperative matching of the operation action and the operation effect cannot be realized, and the quality and the efficiency of the sand suction filling operation are seriously influenced; Fifthly, the compensation for hydrodynamic interference mostly adopts fixed compensation coefficients, the compensation coefficients cannot adapt to the water flow change under all working conditions, the compensation accuracy is insufficient, and the control error is further aggravated. Disclosure of Invention The invention provides a sand sucking and filling multi-arm cooperative control algorithm based on kinematic and dynamic coupling, which improves the bidirectional cooperative optimization of a parasitic-predation algorithm, a fractional order pseudo-spectrum method and a port controlled Hamilton fractional order passive control three-core algorithm through kinematic and dynamic deep coupling modeling, and solves the problem of deep fusion of modeling decoupling, serial flow, weak disturbance resistance and poor cooperative core pain points in the prior art, thereby realizing high-precision, high-stability and high-efficiency cooperative control of multi-arm sand sucking and filling operation in a dynamic water area environment. In order to achieve the above purpose, the invention adopts the following technical scheme: the sand sucking and filling multi-arm cooperative control algorithm based on the kinematic coupling comprises the following steps: S1, acquiring all-condition hydrodynamic interference data of a sand sucking and filling operation water area, and constructing a hydrodynamic interference field feature library after preprocessing the acquired all-condition hydrodynamic interference data; S2, constructing a single mechanical arm kinematics model for each mechanical arm in the multi-arm collaborative operation system, and simultaneously constructing a homogeneous transformation mapping relation between a base coord