73 results about "Barrier function" patented technology

The invention discloses a ring main unit intelligent regulation and control method based on deep reinforcement learning, and the method comprises the following steps: building an electrical topological graph model through collecting the real-time operation data of a ring main unit and a power distribution network, extracting the characteristics of nodes and branches, and generating an initial state vector; a comparison strategy embedding layer is arranged in front of the strategy network, and state-action representation is optimized through comparison learning; constructing a reinforcement learning decision structure by adopting a trust region strategy optimization algorithm, and outputting a ring main unit control action sequence; before execution, an out-of-limit action is judged and corrected by a control barrier function convoy layer; and closing and opening control is completed according to the corrected action, rewards are recorded, and incremental learning and parameter fine adjustment are performed by using an experience playback buffer area. According to the invention, by fusing the graph neural network and deep reinforcement learning and introducing a comparison strategy embedding layer and a control barrier function escorting layer, safe, adaptive and efficient intelligent regulation and control of the ring main unit in a complex power distribution network are realized.

The invention relates to the technical field of control and information, and particularly discloses a distributed optimization control method for a security constraint uncertain multi-agent system, which comprises the following steps of: constructing a topological graph according to a network structure of the multi-agent system, and determining an adjacent matrix of the topological graph; determining a state equation; determining a target function needing to be optimized and inequality constraints needing to be met; according to the state equation, the target function needing to be optimized and the inequality constraint needing to be met, based on a Lyapunov function control method, an optimal solution search condition is established, and based on a barrier function control method, a system safety maintenance condition is established; determining an input constraint condition according to the multi-agent system; calculating by using a quadratic programming method to obtain optimal control input; and controlling the multi-agent system according to the optimal control input. According to the method, the long-term reliability of the system can be remarkably improved, the communication and calculation overhead is remarkably reduced, and the calculation efficiency is remarkably improved.

The invention discloses a hybrid robot adaptive motion control method and system, and relates to the technical field of robot motion control, and the method comprises the steps: obtaining the initial state of a robot system; establishing a nominal dynamics prediction model, and defining a safety set and a control barrier function set; statistical characteristics of model prediction residuals are obtained through online learning; correcting the dynamic model and generating a feed-forward compensation amount, and adaptively adjusting a safety set and a barrier function according to uncertainty; constructing a rolling optimization problem integrating the closed chain constraint and the barrier function hard constraint; solving the optimization problem to obtain an expected control quantity, and adjusting a safety parameter to obtain a rollback control quantity when solving fails; and synthesizing the feed-forward compensation quantity and the control quantity to generate a driving instruction. According to the method, the problem of model mismatching is solved through online learning and residual compensation, safety self-adaption is achieved through uncertainty, performance and safety are considered through a unified optimization framework, and the adaptability, precision and robustness of the system under dynamic disturbance are remarkably improved.

The invention relates to the technical field of power system communication, and particularly discloses an optimization method for an operation and maintenance strategy of a power communication network, which performs space-time alignment and feature reconstruction on multi-source heterogeneous telemetry data of the power communication network by using a sliding window and a graph neural network, and constructs a state tensor reflecting network physical and logic panorama. Generating a candidate strategy set through a deep reinforcement learning action network, and mapping the candidate strategy set to a digital twin environment synchronized with the current network in real time for parallel simulation rehearsal; a dynamic scoring mechanism including network pressure sensing and marginal utility analysis is introduced, fusing is carried out on a high-risk strategy in combination with a logarithmic barrier function, and an optimal strategy giving consideration to performance and safety is selected from a candidate set. And finally, a final execution instruction is generated through verification of a power service hard rule filter, the executed real network is fed back and stored in an experience playback buffer area to update model parameters, and the problems that a traditional static strategy cannot dynamically adapt to the network pressure and the direct trial and error risk is high are solved.

The invention discloses a crane position control method and system based on MPC and a control obstacle function. The method comprises the steps that S101, all to-be-controlled parameter values are calculated according to given positions; s102, solving an optimal control sequence u at the moment according to a predictive control algorithm; s103, designing a state observer according to the output y, the input u and the state at the moment, observing the state at the next moment, designing by adopting a full-dimensional linear state observer method, and adjusting parameters by adopting a DLQR method to obtain state information at the next moment for the step S102 at the next moment; and S104, the output of the real system is measured and sent to the step S103 at the next moment, and the predictive control algorithm is a model predictive control algorithm fused with the control barrier function. According to the crane position control method and system based on the MPC and the control barrier function provided by the invention, the model prediction control is combined with the control barrier function, so that the crane can reach a specified position.

The invention discloses an Internet of Vehicles cooperative driving control method and system based on deep reinforcement learning, and belongs to the field of automatic drive.The method comprises the steps that state data of a main vehicle and state data of at least one adjacent vehicle received through Internet of Vehicles communication are received, generating smooth state estimation and uncertainty characteristics used for representing data credibility; constructing the main vehicle and the adjacent vehicles into a dynamic vehicle interaction diagram; inputting the dynamic vehicle interaction graph into a graph attention feature extraction model to extract space-time interaction features; inputting the space-time interaction characteristics into a deep reinforcement learning strategy network to generate a preliminary reference control instruction; a quadratic programming problem is constructed, and the quadratic programming problem takes minimization of deviation from a reference control instruction as an optimization target and takes security constraints derived from a control barrier function as constraint conditions; and solving the quadratic programming problem to obtain a final control instruction. According to the invention, efficient and intelligent cooperative driving can be realized on the premise of ensuring safety.

This invention relates to the fields of control and information technology, specifically disclosing a distributed optimization control method for multi-agent systems with uncertain security constraints. The method includes the following steps: constructing a topology graph based on the network structure of the multi-agent system and determining the adjacency matrix of the topology graph; determining the state equation; determining the objective function to be optimized and the inequality constraints to be satisfied; establishing optimal solution search conditions based on the control Lyapunov function method and the control barrier function method based on the state equation, the objective function to be optimized, and the inequality constraints to be satisfied; determining the system security maintenance conditions based on the multi-agent system; calculating the optimal control input using a quadratic programming method; and controlling the multi-agent system based on the optimal control input. This invention can significantly improve the long-term reliability of the system and significantly reduce communication and computational overhead, resulting in a significant improvement in computational efficiency.

The present application relates to the technical field of smart water affairs and urban public safety, and particularly relates to a kind of artificial intelligence driven urban pipe network flood resilience evaluation method and system, comprising: constructing Riemann metric tensor by using metric matrix containing potential barrier function, and establishing the Riemann manifold embedding space of pipe network physical state;Then, based on fluid energy, a Hamilton function is constructed, and a dynamics prediction model with physical conservation is obtained by using symplectic neural network and symplectic discrete integral format training;Subsequently, a second-order Hamilton-Jacobi-Aleksandrov partial differential equation describing stochastic differential game is constructed, and a physical perception neural network is used to solve and extract the resilience safety boundary;Finally, real-time data is mapped to the manifold space, the Riemann gradient is calculated, and a quadratic programming problem is solved to generate control instructions. The problem of lack of physical constraints and safety bottom line in pipe network control under extreme random working conditions is solved, and real-time closed-loop control with physical consistency is realized.

This invention relates to the field of robot motion control technology, and more particularly to an obstacle avoidance adaptive dynamic control system, comprising a multimodal environment and physiological temporal perception module with communication connections, a photodynamic Riemannian metric tensor mapping module, a neural constant differential cognitive evolution prediction module, a robust forward invariant tube synthesis module, a Riemannian space barrier constraint solving module, and a differential manifold impedance execution module. The perception module extracts circadian rhythm representation data; the mapping module reconstructs Euclidean space into a Riemannian manifold space based on the photodynamic cost scalar field; the neural constant differential module calculates the cognitive delay evolution trajectory by combining multi-source features; the synthesis module synthesizes a robust forward invariant tube for envelope prediction error; a Riemannian control barrier function constraint proposition is constructed by combining geodesic distance; and the execution module solves for the optimal safe control vector and inversely maps it to impedance torque distribution. This invention achieves deep coupling between physiological rhythms and non-Euclidean geometric control, constructing a physically compliant obstacle avoidance safety defense line.

This invention discloses a safe navigation and localization method for bipedal robots based on discrete control barrier functions and a hierarchical architecture, belonging to the field of robot control technology. The method employs the Relative Dynamics Test (RRT) algorithm to generate a global path and combines it with a Linear Inverted Pendulum Model (LIPM) to generate a stable gait. A discrete control barrier function (D-CBF) is used to ensure obstacle avoidance in dynamic environments, and the safe control input is solved through an optimization problem. Sensor fusion technology is used for real-time localization, and moving average filtering is applied to eliminate oscillation errors caused by gait. This invention, by combining discrete control barrier functions (D-CBF) with hybrid zero dynamics (HZD), achieves high-precision navigation and safe obstacle avoidance for bipedal robots in complex environments, effectively solving the problems of insufficient stability and lack of safety in traditional methods.

The invention provides a power distribution network reactive voltage optimization method based on mathematical analysis, and relates to the technical field of power distribution network control, and the method comprises the steps: S1, building a control model of a distributed active power distribution network reactive power compensation device according to a physical model and control characteristics of the distributed active power distribution network reactive power compensation device; s2, establishing a multi-optimization objective function which takes the minimum network loss as a core and takes voltage quality into account, and establishing operation constraint conditions; s3, based on the obstacle function conversion constraint and gradient / Hessian matrix optimization, designing a solving algorithm model; and S4, integrating the control model, the multi-objective optimization function and the operation constraint condition into an active power distribution network reactive voltage coordination control optimization model, calling a solution algorithm model for solution, and outputting an optimization control scheme of the reactive power compensation device. According to the scheme, the problems that a traditional distribution network optimization model does not fully fuse the characteristics of the distributed active reactive power compensation device, and a multi-dimensional target and complete constraint are not designed for the active power distribution network can be solved.

The present disclosure proposes an actuator failure compensation and safety control method based on a double-layer evaluation architecture, a storage medium and a system. The bottom layer uses an adaptive model predictive control to output a nominal instruction, a reinforcement learning model is built thereon to output a high-frequency residual compensation instruction to eliminate model mismatch, and a second evaluation network is independently built to evaluate the future global loss-of-control risk of the vehicle. The Lagrange dual optimization is introduced to convert the macro risk evaluation into a mathematical constraint to guide the first strategy network correction; the global risk index is used to feed forward the target cost weight of the predictive control to realize continuous flexible degradation, and the control barrier function is triggered at the limit boundary to realize discrete quadratic programming instruction projection. The present application overcomes the conservative strategy degradation defect caused by the traditional mixed reward, establishes an absolute safety line, and improves the control accuracy and safety robustness of the vehicle in complex nonlinear working conditions.

The invention provides a risk-adaptive safe navigation method and device, a terminal and a storage medium, and belongs to the technical field of robot control, and the method comprises the following steps: generating an environment state vector at a current moment; obtaining a risk assessment value; adjusting a security constraint area of the control barrier function and weight distribution of a risk weighted cost function in the model prediction control module; performing optimization solution based on the adjusted risk weighted cost function and the safety constraint formed by the adjusted safety constraint region by using a model prediction control module to obtain a target control instruction meeting the safety constraint and the optimal cost; the robot is driven to move. According to the method, the environment risk is dynamically modeled, the weights of the safety constraint condition and the risk cost function are adaptively adjusted, the adjusted safety constraint condition and the risk cost function are introduced into the model prediction control module to generate the target control instruction, the method can dynamically adapt to the change of the environment, and then the accuracy of navigation is ensured.

The application relates to a kind of nonlinear system safety-relevant control method and system based on reinforcement learning, method includes the following steps: the affine nonlinear dynamic model with safety constraint is established to target nonlinear system;According to affine nonlinear dynamic model, the safety guarantee control input is calculated based on Liapunov type control barrier function;Unknown system dynamics of target nonlinear system is estimated using neural network identifier, and the estimated value of unknown system dynamics is obtained;The nominal tracking control input is calculated according to the estimated value of unknown system dynamics using reinforcement learning architecture composed of actuator and judge;Safety guarantee control input and nominal tracking control input are combined to generate composite control strategy, and the approximate optimal trajectory tracking of target nonlinear system under safety constraint is realized.Compared with prior art, the comprehensive performance and reliability of safety-relevant control in complex nonlinear system are significantly improved.

The invention discloses a self-adjusting predetermined performance consistency control method for a multi-agent system, and the method comprises the steps: building a mathematical model of the multi-agent system, and defining a synchronization error of a follower agent; describing an information interaction relationship among the agents in the multi-agent system based on a graph theory; defining a performance constraint of a synchronization error and a self-adjusting performance boundary function, and introducing a boundary adjusting mechanism with a safety interval detection function to dynamically adapt to disturbance; mapping the synchronization error constrained by the predetermined performance boundary into an unconstrained variable through error conversion and a barrier function; and on the basis of a backstepping method, designing a virtual controller of each order, a final actual controller and an adaptive rate, and realizing cooperative tracking control of the multi-agent system. According to the method, the problems of initial condition feasibility, fixed performance boundary rigid constraint and symmetric performance limitation can be solved, burst interference can be dealt with through a self-regulation mechanism while overshoot, convergence time and steady-state precision are ensured, and the robustness of the system is remarkably improved.

The invention relates to the technical field of robot motion control, and discloses a robot control method and system combining singular point avoidance and obstacle function control. The robot control method is applied to robot control equipment and specifically comprises the following steps that S101, a robot control request is received, a position coordinate system in a Cartesian space is established at the tail end of a mechanical arm, and a tail end position vector x = [x, y, z] T and a virtual wall center coordinate C are defined, the radial distance r = x-c from the tail end to the center of the virtual wall and a radial unit vector are calculated, boundary parameters are set, a virtual wall area is defined by an outer boundary radius rmax and an inner boundary radius rmin, and a safe motion range is formed; s102, decomposing the expected velocity vdes in the Cartesian space into a radial component and a tangential component; the limitation that a traditional potential field method is prone to oscillation and a speed truncation method is discontinuous is broken through by fusing a control obstacle function and a singular point evasion strategy, and strict mathematical constraints established by the control obstacle function ensure that the tail end does not break through a virtual wall boundary absolutely.

The invention discloses a low-altitude pollination control method based on a closed-loop multi-target enhanced optimization algorithm, and aims to solve the problem that the pollination deposition amount is difficult to estimate and feed back in real time. The method comprises the following steps: constructing aerodynamic bias attention, carrying out deposition and uncertainty estimation by adopting a space-time attention network, carrying out reinforcement learning in combination with safety constraints based on a differentiable barrier function and a DreamerV3 world model to generate candidate actions, and refining the actions under constraint guidance by using a diffusion model to form closed-loop control. The technical effects that the deposition uniformity is improved, non-target drift is reduced, mechanical impact and energy consumption of flowers are restrained, and the safety distance and spraying forbidding area constraint are met are achieved.

The invention relates to an MIMO-DFRC waveform design method based on angle estimation CRB optimization. The method comprises the following steps: establishing a signal model; deriving an angle estimation CRB; establishing a communication CI constraint; establishing a constant modulus constraint; an MIMO radar communication integrated waveform design problem based on angle estimation CRB optimization is established; converting inequality constraints by adopting a barrier function thought; and a Riemann conjugate gradient method is adopted for efficient solution. According to the method, the thought of a barrier function in an interior point method is adopted to process inequality constraints, it is ensured that the optimization result can strictly meet the established communication service quality requirement, and the method is more suitable and robust in an actual radar communication integrated scene; according to the method, the Riemann conjugate gradient method is adopted to carry out integrated waveform optimization solution, and compared with an existing ADMM algorithm, the calculation efficiency is higher, the real-time emission capability of MIMO radar communication integrated waveforms is improved, and the method has more advantages in engineering application.

The invention specifically discloses a software and hardware collaborative multi-unmanned aerial vehicle formation control method and system. The method comprises the following steps: collecting state information and a neighborhood observation set of each unmanned aerial vehicle; outputting a reference control quantity by using a reference strategy generator; screening neighborhood entities needing security constraint from the neighborhood observation set to form a local graph structure; setting a graph structure safety projection unit, constructing a barrier function based on neighborhood topology, converting the barrier function into a linear inequality constraint for control input, assembling the linear inequality constraint into a matrix inequality, solving a quadratic programming problem, and obtaining a final control quantity meeting the constraint; and carrying out amplitude limiting and rate limiting on the final control quantity or mapping the final control quantity into an expected instruction expected by flight control, and issuing the flight control instruction through an agreed protocol. According to the technical scheme, based on software and hardware cooperation, a high-level reference control strategy and real-time safety correction based on a local graph structure are organically combined, and the deterministic safety delay boundary and the fault isolation capacity are obtained while the control performance is guaranteed.

Aiming at the problem of missile terminal guidance with no-fly zone and impact angle constraints, a model predictive static programming guidance law based on control barrier function is proposed. The proposed algorithm inherits the high efficiency of the traditional model predictive static programming while ensuring the no-fly zone constraints. The main reasons are as follows: first, the projection operator under the linear inequality constraints has an explicit expression, avoiding additional optimization solving; second, due to the forward invariance of the control barrier function, the no-fly zone constraints do not need to be imposed at all discrete nodes, but only at one or a few discrete nodes, reducing the scale of solving. The simulation results show that the proposed algorithm can meet the no-fly zone and impact angle constraints while ensuring high computational efficiency. Subsequent research will focus on trajectory optimization and guidance problems with multiple no-fly zone constraints, even dynamic no-fly zone constraints.

The application discloses a low-altitude pollination control method based on a closed-loop multi-target reinforcement optimization algorithm, in order to solve the problem that pollination deposition is difficult to estimate and feedback in real time, local aerodynamics and particle migration field are obtained through multi-source sensing and Fourier neural operators, aerodynamic bias attention is constructed, and deposition and uncertainty estimation are carried out by using a space-time attention network, candidate actions are generated by combining safe constraints based on a differentiable barrier function and DreamerV3 world model reinforcement learning, and the diffusion model is used to refine the action under the guidance of constraints to form a closed-loop control, so that the technical effects of improving deposition uniformity, reducing non-target drift, inhibiting flower mechanical impact and energy consumption, and meeting the safety distance and no-spraying area constraints are achieved.

The invention provides a man-machine sharing control-oriented dynamic permission allocation system and method. The method comprises the following steps of: establishing a vehicle longitudinal-transverse coupling dynamic model; the robust security constraint is generated through an environment robust control barrier function, and the environment robust control barrier function compensates the influence of sensor measurement uncertainty through a residual error correction term; the performance constraint is generated by controlling a Lyapunov function; based on the robust security constraint and the performance constraint, the change rate of the control permission of the automatic system is used as a core decision variable, an optimization problem is constructed, the robust security constraint is set as a hard constraint, the performance constraint is set as a soft constraint, and the optimization objective is to minimize the permission fluctuation degree and the control error at the same time; and solving the optimization problem to obtain a current control authority weight, and carrying out weighted fusion on the driver control instruction and the automatic system control instruction to generate a final vehicle control instruction.

The application discloses a kind of MHC rotary crane trajectory planning method and system based on safety barrier function, method includes: the dynamic model of MHC rotary crane is established, based on dynamic model, the rotation angle of MHC rotary crane and rope length are solved out by the position information of port crane;Design LQR-CSCBF controller, control the system after linearization using LQR to obtain nominal control quantity, construct safety constraint condition using CSCBF, calculate the real control quantity that satisfies anti-sway safety constraint;LQR-CSCBF-RRT* algorithm is used for trajectory planning, random node is generated by sampling in configuration space, the trajectory from parent node to child node is simulated using LQR-CSCBF controller, find the optimal trajectory from start point to end point without collision and satisfy anti-sway constraint.The MHC rotary crane trajectory planning method and system based on safety barrier function provided in the application, trajectory optimization is carried out by LQR-CSCBF-RRT* algorithm, output a piece of obstacle avoidance constraint and safety constraint, so that crane can reach the optimal trajectory from start point to target point.

The application discloses an AC solid-state circuit breaker adaptive turn-off method and system, the core of the method is to construct a two-dimensional phase space geometric evolution manifold according to the current amplitude sequence of the fault current and the hardware differential rate signal, realize geometric extrapolation of the short-circuit current peak value containing the non-periodic component through the trajectory curvature characteristics; combined with the cumulative equivalent operation times, the radiator temperature and the standing time and other physical state parameters, input them into the endogenous physical information neural network model, define the safe operation interval with physical certainty by using the hard-coded embedded Cauer thermal network operator; and in the interval, an analytical projection algorithm based on stress cost manifold is adopted to execute constant level optimization of the optimal turn-off time by using the control barrier function. The application can effectively solve the distortion problem of the traditional model under the magnetic saturation working condition and the uncertainty of the soft constraint prediction, while ensuring the breaking reliability, significantly reducing the online calculation complexity.

The present application relates to a kind of based on online quality monitoring's flotation optimization control method, including time acquisition the data collected in the multiple links of entire flotation process, the data set is obtained by processing the data collected, the causal feature of data set is extracted using convergence cross mapping method, the dynamic relationship model of key operation variable and production index variable is obtained;Design the optimization objective of optimization control based on Lyapunovo Barrier function to realize single-step;Optimal action is searched using time forward rolling type finite time domain optimization strategy, and the optimization control task of cycle forward is realized;Step 8, repeat step 5, 6, 7, and the flotation process is controlled by online optimization.The advantages of the present application are: effectively avoid the disconnection of fixed global optimization target and actual production, and the target variable is reasonably optimized, such as reducing the grade of float tailings, and the grade of float fine product is stable, which has better effect.

The application relates to a MIMO-DFRC waveform design method based on angle estimation CRB optimization, and the method comprises the following steps: establishing a signal model; deducing angle estimation CRB; establishing a communication CI constraint; establishing a constant modulus constraint; establishing a MIMO radar communication integrated waveform design problem based on angle estimation CRB optimization; converting inequality constraints by adopting the idea of barrier function; and efficiently solving by adopting the Riemann conjugate gradient method. The method adopts the idea of barrier function in the interior point method to process inequality constraints, ensures that the optimization result can strictly meet the established communication service quality requirement, is more applicable and stable in the actual radar communication integrated scene, adopts the Riemann conjugate gradient method to solve the integrated waveform optimization, is higher in calculation efficiency compared with the existing ADMM algorithm, improves the real-time transmission capability of the MIMO radar communication integrated waveform, and is more advantageous in engineering application.

This invention relates to the field of power grid inspection technology, and more particularly to a spatiotemporal dynamic path planning method for power grids integrating GNN-UTVD-CBF. This method first constructs a hybrid power grid environment model and calculates the safe time intervals for potential path edges based on equipment dynamics constraints. Then, it uses a graph neural network to predict edge expansion priorities and combines a unified temporal visibility deformation criterion to construct a dynamically connected visibility map, generating a candidate path set with multiple topological features. In path search, a safety verification mechanism based on a linear quadratic regulator and a high-order control barrier function is integrated to achieve efficient safety verification of the quadratic programming solution. Finally, a smooth trajectory is generated through spatial-temporal corridor expansion and B-spline curve optimization. When the environment changes dynamically, the system quickly adjusts the path through local pruning and incremental repair algorithms. This invention significantly improves the planning efficiency, safety, and robustness of power grid inspection equipment in complex dynamic environments.

The present disclosure relates to reusable barriers for synchronization between multiple processes. In one aspect, a system and method for facilitating synchronization between processes is provided. During operation, the system can execute multiple processes in parallel on one or more computing nodes. In response to a first process calling a barrier function, the system can suspend execution of the first process, and in response to determining that the first process has gained access to a variable shared by at least a subset of the multiple processes, the system can update the shared variable. The system can release the shared variable to a second process in the subset to update the shared variable when the second process calls the barrier function, and determine whether all processes in the subset have updated the shared variable. In response to the shared variable having been updated by all processes in the subset, the system can resume execution of all processes in the subset.

The invention provides a multi-agent formation control method and device based on performance and security decoupling control, electronic equipment and a readable storage medium. According to the method, a nominal controller meeting performance constraints is designed based on PPC; and the nominal controller adopts a time-varying performance function with an auxiliary variable and is used for limiting the tracking error to evolve within a predefined dynamic limit. And designing a control barrier function (CBF) as a safety constraint, taking an intelligent agent control input smaller than a known scalar as an input saturation constraint condition, modeling a quadratic programming problem, and solving a safety control quantity closest to a nominal control quantity output by a nominal controller. And adjusting an auxiliary variable of the time-varying performance function according to the difference of the sum so as to relieve the conflict between the performance and the security and eliminate the control singularity. The multi-agent formation control method can solve the problem of multi-agent formation control under the condition that performance, safety and input constraints exist at the same time.

The application relates to a kind of multi-mobile robot obstacle avoidance control methods based on dynamic control barrier function, comprising the following steps: S1, obtaining multi-mobile robot state information and dynamic obstacle point cloud data;S2, pre-processing is obtained with several obstacle point cloud clusters;S3, ellipse parameterization fitting is obtained with ellipse obstacle parameters;S4, based on ellipse obstacle parameters, the ellipse obstacle sequence in prediction time domain is obtained;S5, extended ellipse obstacle sequence is obtained;S6, establishment formation reference model generates reference trajectory;S7, construct discrete-time dynamic control barrier function constraint as safety constraint;S8, calculate obstacle avoidance risk index and obstacle avoidance responsibility coefficient and adjust formation keeping weight and obstacle avoidance constraint weight in model predictive control;S9, based on the weight after adjustment, establish model, obtain prediction control input sequence, and carry out rolling optimization.The beneficial effects of the application are: improve modeling accuracy, motion trend utilization rate and prediction error safety margin accuracy.