CN112035992B - Multi-objective optimization-based autonomous remote control underwater robot sharing control method - Google Patents
Multi-objective optimization-based autonomous remote control underwater robot sharing control method Download PDFInfo
- Publication number
- CN112035992B CN112035992B CN201910401020.2A CN201910401020A CN112035992B CN 112035992 B CN112035992 B CN 112035992B CN 201910401020 A CN201910401020 A CN 201910401020A CN 112035992 B CN112035992 B CN 112035992B
- Authority
- CN
- China
- Prior art keywords
- arv
- function
- obstacle
- control command
- heading angle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000005457 optimization Methods 0.000 title claims abstract description 66
- 238000011156 evaluation Methods 0.000 claims description 15
- 238000013461 design Methods 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 8
- 230000003247 decreasing effect Effects 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 7
- 238000011430 maximum method Methods 0.000 claims description 6
- 230000004888 barrier function Effects 0.000 claims 1
- 230000007613 environmental effect Effects 0.000 description 6
- 230000006399 behavior Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000007423 decrease Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to an autonomous remote control underwater robot sharing control method based on multi-objective optimization, which comprises the following steps: according to the input signal of the operating lever, calculating an ARV heading angle control command and a forward speed control command which are manually operated; calculating an ARV heading angle control command of autonomous control; constructing a shared objective function, constraint conditions and an optimization solving method thereof, and outputting a heading angle shared control command according to the current position of the ARV, the heading angle, the distribution position of the obstacle in the local environment, the calculated manually operated ARV heading angle control command and the autonomously controlled ARV heading angle control command; the ARV avoids the obstacle and executes the task according to the control command of the forward speed of the manual operation and the sharing control command of the heading angle. The method combines the remote control and the autonomous control of an operator to improve the working capacity, the task performance and the safety of the ARV in a complex, unknown and unstructured underwater environment, and lightens the workload of the operator.
Description
Technical Field
The invention relates to the field of underwater robot control, in particular to an autonomous remote control underwater robot man-machine sharing control method based on multi-objective optimization.
Background
The autonomous remote control underwater robot (Autonomous and Remotely Operated Underwater Vehicle, ARV) is a novel underwater robot integrating the characteristics of an AUV and an ROV, is self-contained with energy, uses optical fibers to communicate with a mother ship, has the functions of AUV large-scale underwater searching and detecting, and can be operated and controlled in real time by an operator like the ROV to realize fixed-point observation and underwater light operation. ARV provides convenience for human exploration, research and development of the ocean.
The main control modes of the ARV at present comprise autonomous control and remote control. Limited by the state of the art of sensors and intelligent technology, autonomous control is generally used for performing relatively simple tasks such as preprogrammed searching, observation and the like, and has limited application scenarios; when a task is executed in a complex and unknown underwater environment, a remote control mode is generally adopted, an operator directly controls the ARV to observe or work in the remote control mode, the task performance of the ARV system depends on the technical level of the operator, the safety of the ARV is low, and the workload of the operator is heavy.
Disclosure of Invention
Aiming at the problems existing in the autonomous control and the remote control, the invention aims to provide an ARV man-machine sharing control method based on multi-objective optimization, which improves the working capacity, task performance and safety of the ARV in a complex, unknown and unstructured underwater environment by combining the remote control and the autonomous control of an operator, and simultaneously reduces the workload of the operator for remotely controlling the ARV.
The technical scheme adopted for solving the technical problems is as follows: an autonomous remote control underwater robot sharing control method based on multi-objective optimization comprises the following steps:
the current position P of the ARV detected by the operator on the basis of the sensor ARV Heading angle psi ARV And the distribution position P of the obstacle in the local environment obs Operating the operating lever, and calculating an ARV heading angle control command psi of manual operation according to the input of the operating lever h And control command v of forward speed h ；
According to P obs And P ARV Calculating an autonomous ARV heading angle control command ψ apf ；
Constructing a shared objective function, constraint conditions and an optimization solving method thereof, and according to the current position P of the ARV ARV Heading angle psi ARV And the distribution position P of the obstacle in the local environment obs And a calculated manually operated ARV heading angle control command ψ h Autonomous ARV heading angle control command ψ apf Outputting a heading angle sharing control command psi share ；
ARV control command v according to forward speed h Control command psi for sharing heading angle share And controlling the ARV to avoid the obstacle and executing the task.
The ARV heading angle control command psi of manual operation is calculated h And control command v of forward speed h The formula of (2) is as follows:
wherein t is the time; j (J) x And J y For operator control input on the x and y axes of the lever, where J x Control of ARV heading angle, J y Control the forward speed of ARV, and J x ∈[-1，1]、J y ∈[-1，1]；ψ ARV The current heading angle of the ARV; k (k) ψ Gain factors for adjusting the heading angle increment; v max Is the maximum forward speed of the ARV; j (J) xd Not less than 0 and J yd And ≡0 is a dead zone threshold set near the x-and y-axis 0 positions of the lever, respectively.
Said according to P obs And P ARV Calculating ARV heading angle control command psi of autonomous control system apf Comprising the following steps:
(1) if an obstacle exists in the environment, the ARV heading angle control command psi of the autonomous control system is calculated by using the following formula apf ：
ψ apf (t)＝atan(y v -y o ，x v -x o )
Wherein P is ARV ＝[x v ，y v ]For ARV current position coordinates, [ x ] o ，y o ]Is the position of the obstacle;
(2) if there are multiple obstacles in the environment, then:
calculating the magnitude r of the effect of the ith obstacle on ARV motion using the following formula i ：
Wherein i is the number of the obstacle, d obs_i For Euclidean distance, d, between ARV and the ith obstacle share 、d safe 、k o D is a design parameter share >0 is the maximum distance of the obstacle to the ARV navigation safety, d safe >0 is a distance threshold value for guaranteeing safety between ARV and obstacle, d share 、d safe K is determined according to the radius of motion of ARV o (1≥k o >0) The self-obstacle avoidance behavior is a proportionality coefficient and is used for adjusting the degree of assisting other behaviors in obstacle avoidance of the self-obstacle avoidance behavior;
by r i Is the size of the module, ψ apf_i Determining for direction a vector V of the effect of the ith obstacle on ARV obs_i The vector sum of all the obstacle's influence on the ARV is obtained by vector addition using the following equation, using the direction of the vector sum as the control command ψ outputted from the autonomous control module apf ：
Where n is the number of obstacles in the environment that affect the safety of ARV navigation.
The shared objective function comprises a compliance function obedience (psi), an autonomy (psi) and a stability function stability (psi) and is used for evaluating different values of a heading angle control command variable psi in the following forms:
aim(ψ)＝exp(-χ|ψ-ψ aim |)
in the formula, aim represents an objective function, ψ aim Exp () is an exponential function with a natural constant as a base, which is an optimization target of the objective function; x is more than or equal to 0; the value ranges of the objective functions of this form are all in the interval (0, 1]An inner part; the magnitude of the function value of the objective function and the heading angle control command variable psi meet the corresponding objective psi aim Is positive in degree of correlation, at symmetry axis ψ=ψ aim The maximum value 1 is obtained and the function value decreases to two sides, the decreasing speed of the function value can be changed by adjusting the shape coefficient χ, the greater the χ is, the faster the decreasing speed is, and in particular, when χ=0, the function value is constant at 1.
The compliance function obedience (ψ) is used for evaluating the ψ compliance aim The degree of (3):
obedience(ψ)＝exp(-α|ψ-ψ h |)
alpha is the shape factor of the compliance function, the value of which is determined by:
wherein alpha is max A constant greater than 0; d, d mmin Is the minimum distance of the ARV to the obstacle; the formula is according to d mmin The value of alpha is adjusted to change the extent to which the operator affects the ARV motion.
The autonomy (ψ) is used for evaluating the ψ and the ψa pf Is the difference between (a):
autonomy(ψ)＝exp(-γ|ψ-ψ apf |)
gamma is the shape factor of the autonomous function, the value of which is determined by:
wherein, gamma max For constants greater than 0, the formula is according to d min Adjusting the value of gamma to make phi apf At d only min ＜d share The ARV is moved away from the obstacle.
The stability function stability (ψ) is used to evaluate the ratio of ψ to ψ ARV The degree of variation of (2):
stability(ψ)＝exp(-β|ψ-ψ ARV |)
where β is the shape factor of the stability function and is a normal number.
Constraint condition I of the shared objective function safe According to the security evaluation function, the method comprises the following steps of:
firstly, determining a set I 'of value intervals of a motion direction meeting the safety requirement' safe ＝[I′ 1 ，I′ 2 ，...，I′ j ]J is the number of the value intervals, so that ARV is in I' safe In the direction of movement within the interval of (2) andthe obstacle keeps a safe distance; using security (ψ) < λ (d) max -d share ) Calculating to obtain I' safe ；
Wherein the security assessment function is defined as a piecewise function, and the function value on each piece of domain is determined according to the following formula:
where security (ψ) represents a security evaluation function; μ is a scaling factor changing the magnitude of the function value, set to a constant greater than 0; d, d max A distance threshold value that causes a change in the security (ψ) function value for the obstacle to start, andd is the minimum distance from ARV to obstacle grid within a single definition domain, when d>d max Time d=d max ；
Secondly, setting a safety interval I for guaranteeing ARV safety Δ Reduce I' safe To obtain a set I of feasible solution value intervals safe ＝[I 1 ，I 2 ，...，I m ]M is the number of the value intervals, and the ARV is ensured to be in I safe The distance to the obstacle on both sides in the movement direction in the interval of (2) is also kept relatively safe.
The optimal solving method of the shared objective function adopts a minimum maximum method, and a heading angle control command psi generated by the shared control method can be obtained through solving share The following formula is shown:
where min { } represents that each segment in the defined domain takes the minimum of all objective functions.
The invention has the following beneficial effects and advantages:
1. the method of the invention improves the task performance of the ARV in complex, unknown and unstructured environments by combining the shared control with the environmental awareness, decision making capability of operators and the accurate control capability of an autonomous control system.
2. The method of the invention uses the control signal generated by the autonomous control system to assist the operation of the operator, thereby reducing the operation complexity of the operator in the task and reducing the workload.
3. The method designs a method for processing the local environment information acquired by the ARV sensor, and adjusts the control signal of the ARV according to the information obtained after the processing, thereby ensuring the safety of the ARV.
4. The method can give consideration to various factors such as task demands, operator demands and the like when integrating the remote control of the operators and the autonomous control of the autonomous control system, is suitable for various tasks, and has wide application scenes.
5. The method adopts a modularized design thought, has simple structure and strong expandability, and is convenient for adjusting the design of the algorithm according to task requirements.
6. The method is based on the existing multi-objective optimization theory design, can be compatible with a large number of mature objective function forms and optimization algorithms in the multi-objective optimization theory, and can meet wide actual demands.
Drawings
FIG. 1 is a system block diagram of the method of the present invention;
FIG. 2 is a block diagram of the method of the present invention;
FIG. 3 is a schematic illustration of processing local environmental information acquired by an ARV sensor in the method of the present invention;
FIG. 4 is a schematic representation of a solution using a least squares method in the method of the present invention;
figure 5 is a flow chart of the method steps of the invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples.
The basic idea of the method is that based on the theory of sharing control, a control command of ARV heading angle is used as a decision variable, a group of corresponding objective functions are designed according to factors such as control intention of operators, autonomous control, task requirements, requirements on ARV task performance and the like, constraint conditions of ARV in the tasks are determined, the sharing control of the ARV is converted into a multi-objective optimization problem by using the objective functions and the constraint conditions, and an appropriate optimization algorithm is selected in the multi-objective optimization method to solve the problem so as to obtain an optimal solution (namely a Pareto optimal solution in multi-objective optimization) as the heading angle control command generated by ARV sharing control. The method can give consideration to various factors in the sharing control of the ARV, improves the operation capability and the safety of the ARV in complex, unknown and unstructured underwater environments, and simultaneously reduces the workload of operators.
Based on the method of the invention, the embodiment aims at the problem that the ARV performs environment exploration task design in the underwater environment with unknown global environment information and unstructured, the moving target and the moving path of the ARV cannot be preset in the task, and an operator is required to search, decide and plan by means of real-time local environment information acquired by acoustic, optical and other sensors carried by the ARV, the safety is poor, the task performance is limited by the technical level of the operator, the workload of the operator is heavy and the like, and the shared control method based on multi-target optimization designed by the embodiment can well solve the problem.
P in FIG. 1 obs ＝[P obs_1 ，P obs_2 ，...，P obs_n ](n is the number of obstacles) is the position information of the obstacle distribution in the local environment, P ARV For the current position of ARV, ψ ARV Is the current heading angle of ARV, ψ h ARV heading angle control command, v, for operator output h Control command of ARV forward speed, ψ, for operator output apf Control command of ARV heading angle outputted by autonomous control system, psi share The control command is shared for the heading angle generated by using the sharing control method based on multi-objective optimization. In FIG. 3, small squares resulting from the intersection of the dashed lines are grids, where black grids represent obstacle grids, white grids represent free grids, Δ represents grid granularity, Δ is determined from the accuracy of the local obstacle information and the performance of the computer, and smaller Δ is for computer performance and map accuracyThe higher the demand of (2); the large square marked by the solid line is an active window, w is the side length of the active window, and the value of w is determined according to the detection distance of the sensor; the ellipse represents an ARV, the center of gravity of which is located in the grid of the center of the movable window; θ is the angle used to divide the active window, and a group of rays centered on the ARV location are used to divide the active window intoDetermining the size of theta according to the performance of a computer and the precision of a local map, wherein the smaller the theta, the higher the requirements of the computer on the performance and the precision of the map are, and the more accurate the obtained security evaluation function describes the environment; d is the minimum distance from ARV to obstacle grid in a single area, let +.>I′ 1 To meet the security requirement, i.e. the value interval of the psi of the security evaluation function, I 1 For constraint I safe A value interval of the medium psi; i Δ For safety interval, I Δ Is determined according to the requirements of ARV safety in the task, I Δ The greater the ARV the greater the safety of the ARV, but the corresponding decrease in the range of motion of the ARV near the obstacle. In FIG. 4, I 1 、I 2 And I 3 Respectively is I safe In the value interval of the heading angle control command variable psi, the obedience (psi) is a compliance function in an objective function, the autonomy (psi) is an autonomy function in the objective function, the stability (psi) is a stability function in the objective function, and the black bold line is a single-objective optimization function min { obedience (psi), autonomy (psi), and stability (psi) meeting constraint conditions, which are obtained when a minimum maximum method is used for calculation.
As shown in fig. 1, in this embodiment, a modularized system structure control ARV is designed, and is divided into four modules, namely a sensing system module, an operator remote control module, an autonomous control module and a shared control module.
The sensing system module is used for acquiring and outputting motion states of the environment and ARVStatus-related information including distribution P of obstacles in local environment obs ＝[P obs_1 ，P obs_2 ，...，P obs_n ](n is the number of obstacles), the current position P of ARV ARV And a heading angle psi ARV . The sensor that adopts can detect environmental information, motion state and positional information, includes: acoustic, optical, etc. sensors, such as forward looking sonar, anti-collision sonar, cameras, etc.
The function of the remote control module of the operator is P output by the operator according to the sensing system module ARV 、ψ ARV And P obs The ARV heading angle control command psi of the operator is respectively output through the x axis and the y axis of the remote control operation lever h And control command v of forward speed h ，ψ h And v h Calculated as follows:
wherein t is the time calculated by using the formula; j (J) x And J y For operator control input on the x and y axes of the lever, where J x Control of ARV heading angle, J y Control the forward speed of ARV, and J x ∈[-1，1]、J y ∈[-1，1]；ψ ARV The current heading angle of the ARV; k (k) ψ Determining a gain coefficient for adjusting the increment of the heading angle according to the control capability of the ARV motion controller and the steering capability of the ARV; v max Is the maximum forward speed of the ARV; j (J) xd Not less than 0 and J yd And the dead zone threshold value is set near the x-axis and the y-axis 0 positions of the operating rod respectively, so that an operator can conveniently input a control command for keeping the current moving direction and the forward speed of the ARV to be 0, and the control command is determined according to the technical level of the operator and the requirement of a task on the control command of the operator.
The autonomous control module outputs P according to the sensing system module obs And P ARV The ARV heading angle control command ψ of the autonomous control system is calculated using the following formula apf ：
ψ apf (t)＝atan(y v -y o ，x v -x o )
Wherein P is ARV ＝[x v ，y v ]For ARV current position coordinates, [ x ] o ，y o ]As the position of the obstacle, atan (·) is an arctangent function.
If there are multiple obstacles in the environment, the size r of the effect of the ith obstacle on ARV motion is calculated using the formula i (i is obstacle number):
wherein d obs_i For Euclidean distance, d, between ARV and the ith obstacle share 、d safe 、k o D is a design parameter share > 0 is the maximum distance that the obstacle has to affect on ARV navigation safety, d safe > 0 is the distance threshold value, d, for ensuring safety between ARV and obstacle share 、d safe K is determined according to the radius of motion of ARV o (1≥k o > 0) is a proportionality coefficient for adjusting the degree of the autonomous obstacle avoidance behavior to assist other behaviors in obstacle avoidance, and determining according to the technical level of operators and the requirement on safety, wherein the higher the requirement on safety is, the k is o The larger. By r i Is the size of the module, ψ apf_i Determining for direction a vector V of the effect of the ith obstacle on ARV obs_i The vector sum of all the obstacle's influence on the ARV is obtained by vector addition using the following equation, using the direction of the vector sum as the control command ψ outputted from the autonomous control module apf ：
Where n is the number of obstacles in the environment that affect the safety of ARV navigation.
The sharing control module outputs psi according to the sensing system module ARV 、P ARV And P obs Psi output by remote control module of operator h Psi outputted by the autonomous control module apf Generating a shared control command psi of a heading angle using a shared control method based on multi-objective optimization share 。
In the configuration shown in FIG. 1, the forward speed of the ARV is controlled by the operator output control command v h Direct control, shared control command psi generated by shared control module by using shared control method based on multi-objective optimization share And (5) controlling.
As shown in fig. 2, the autonomous control module in fig. 1 is configured to generate a heading angle sharing control command ψ share The sharing control method based on multi-objective optimization comprises three parts of an objective function, constraint conditions and an optimization method, wherein the constraint conditions are determined by a security evaluation function. The objective function and constraint condition transform the shared control of ARV heading angle into control command variable psi (psi E (-pi, pi)]) As a multi-objective optimization problem of decision variables, solving the problem by an optimization method to obtain the optimal solution of psi, namely the shared control command psi of ARV heading angle share . The objective function, constraint and optimization method are described in detail below.
In the embodiment, three objective functions of obedience (psi), autonomy (psi) and stability (psi) are respectively designed to evaluate different values of psi with the aim of obeying the control intention of operators, improving the safety of ARVs, reducing the operation complexity of operators and optimizing the motion path of ARVs. The objective functions are all designed into a negative exponential function form based on natural constants, and the negative exponential function is shown as the following formula:
aim(ψ)＝exp(-χ|ψ-ψ aim |)
in the formula, aim represents an objective function, ψ aim Exp () is an exponential function with a natural constant as a base, which is an optimization target of the objective function; x is more than or equal to 0; the value ranges of the objective functions of this form are all in the interval (0, 1]In, facilitate the adaptation to different objective functionsIs compared with the function value of (a). The size and psi of the function value of the objective function meet the corresponding objective psi aim Is positive in degree of correlation, ψ=ψ at the symmetry axis aim The maximum value 1 is obtained and is decreased to two sides, the decreasing speed of the function value can be changed by adjusting the shape coefficient χ, the greater the χ is, the faster the decreasing speed is, and particularly, when χ is zero, the function value is constant to be 1; in this embodiment, the minimum-maximum method is used as an optimization method to solve the multi-objective optimization problem, and the optimization result is determined by the objective function with the minimum function value, so that the larger the shape coefficient of the objective function is, the larger the influence on the optimization result is, and when the shape coefficient is 0, the influence on the optimization result is avoided, so that the effect of each objective function in the multi-objective optimization problem solution can be conveniently adjusted by adjusting the shape coefficient. The design of the three objective functions is described below.
In an environmental exploration task, the movement of an ARV first needs to follow the control intent of the operator to meet the operator's observation needs. The embodiment adopts the psi h Representing the control intention of the operator, a compliance function obedience (ψ) shown in the following formula is designed for evaluating the compliance of ψ h The degree of (3):
obedience(ψ)＝exp(-α|ψ-ψ h |)
where α is the shape factor of the compliance function, and its value is determined by:
wherein alpha is max For a constant greater than 0, according to the skill level setting of the operator, the higher the skill level, the α max The larger the ARV is, the more compliant the operator's control commands are in the shared control based on multi-objective optimization; d, d min Is the minimum distance of the ARV to the obstacle; the formula is according to d min Adjusting the value of alpha to change the extent to which the operator affects ARV movement, d min The smaller the shape factor of the compliance function, the smaller the influence of the compliance function on the optimization result, the corresponding decrease, the operator remotely controls the weight decrease in the sharing control of the ARV,the ARV is more subject to control commands of the autonomous control system to secure its safety.
The ARV should also obey the control command ψ of the autonomous control system in the environment exploration task apf To assist the operation of the operator and improve the safety of the ARV, the present embodiment designs an autonomy (ψ) as an autonomy function for evaluating ψ and ψ apf Is a difference in (a) between the two.
autonomy(ψ)＝exp(-γ|ψ-ψ apf |)
Where γ is the shape factor of the autonomy function, and its value is determined by:
wherein, gamma max A constant greater than 0. The formula is according to d mmin Adjusting the value of gamma to make phi apf At d only min ＜d share The ARV is moved away from the obstacle, and the influence of the shape factor on the objective function in the multi-objective optimization can be determined, wherein when the ARV is far away from the obstacle, the autonomous control system has no influence on the movement of the ARV.
The appropriate reduction of the abrupt change of the control command with respect to the current movement direction may make the movement state of the ARV more stable, thereby reducing the complexity of the operator's operation and making the movement path of the ARV smoother. Therefore, the present embodiment designs stability function stability (ψ) shown in the following formula for evaluating the relationship of ψ with respect to ψ ARV The degree of variation of (2):
stability(ψ)＝exp(-β|ψ-ψ ARV |)
wherein beta is>0 is the shape coefficient of the stability function, the larger the beta is, the larger the influence of the stability function on the optimization result is, and the generated psi is optimized share Relative to psi ARV The smaller the variation of (c). Too large β will make ARV less prone to change direction of motion, thus setting β to a small positive constant.
As shown in fig. 3, the constraint condition is a set I of the value intervals of the heading angle control command variable ψ safe ＝[I 1 ，I 2 ，...，I m ](m is the number of the value intervals) and is determined by a security evaluation function; the safety evaluation function is calculated according to the distribution of the obstacles in the local environment so as to ensure the safety of the ARV in the task. In the environment exploration task, the present embodiment determines a security evaluation function according to the principle that the greater the distance between the ARV and the obstacle, the higher the security, the direction and distance information of surrounding obstacle distribution acquired by acoustic, optical, etc. sensors carried by the ARV. The grid map within the active window of fig. 3 is used to calculate a security assessment function.
In the present embodiment, the security evaluation function is designed as a piecewise function, the independent variable of which is ψ (i.e. the definition domain is the interval (-pi, pi ]). In the polar coordinate system, the definition domain is piecewise according to the region division with the included angle θ in fig. 3, the function value in each definition domain is constant, the size of the function value is determined according to d in the corresponding region in fig. 3, the larger d is, the smaller the function value is, the safer the value of ψ in the definition domain of the segment is, and the function value on each definition domain is calculated by using the following formula:
where security (ψ) represents a security evaluation function; μ is a scaling factor changing the magnitude of the function value, set to a constant greater than 0; d, d max A distance threshold value that causes a change in the security (ψ) function value for the obstacle to start, andd is the minimum distance from ARV to the obstacle grid in a single defined area, let it be when no obstacle grid is present in the corresponding area
Constraint I in this embodiment safe According to the safety evaluation function, two steps of calculation are carried out, firstly, a set I 'of value intervals of the motion direction meeting the safety requirement is determined' safe ＝[I′ 1 ，I′ 2 ，...，I′ j ](j is a value intervalNumber of (2) to make ARV at I' safe Distance to the obstacle kept safe in the movement direction within the interval of (2): use d share As a distance threshold for securing ARV, the corresponding safety evaluation function has a function value threshold of λ (d max -d share ) Since the larger the distance is, the smaller the function value is, and the safer the value of ψ is in the calculation of the function value of the security evaluation function, the security (ψ) < λ (d is used max -d share ) Calculating to obtain I' safe . Then, the I 'is reduced' safe To obtain I safe To ensure ARV at I safe The distance to the obstacle on both sides in the movement direction in the interval of (2) is also kept relatively safe: as shown in FIG. 3, a safety interval I for securing ARV is provided Δ In I' safe All the value intervals (in I' 1 For example) reduces I at each end Δ To obtain I safe (as I) 1 For example).
Using the design of objective functions and constraints described above, the sharing control of ARV heading angles can be translated into a maximized multi-objective optimization problem as shown in the following equation:
where max represents ψ corresponding to the maximum value of the objective function value preferentially taken in the multi-objective optimization.
The optimization method is used for solving the multi-objective optimization problem to obtain an optimal solution psi share . In the environment exploration task, the ARV usually works in an underwater environment with unknown global environment information, the requirement on reliability is very high, and an optimization algorithm should be solved in each control period to obtain a stable and reliable result; furthermore, since the operator needs to control the ARV in real time, the optimization algorithm selected should have less time complexity to reduce the amount of computation. Therefore, the present embodiment uses the minimum maximum method as an optimization method to seek the best optimization result in the worst case. The minimum and maximum method requires less calculation amount, and simultaneously has stable optimization result and can ensureThe real-time performance and reliability of the ARV sharing control method in the task are ensured. The maximized multi-objective optimization problem is converted into a single-objective optimization problem by a minimum maximization method, and a heading angle control command psi generated by a shared control method can be obtained by solving share The following formula is shown:
where min { } represents that each segment in the defined domain takes the minimum of all objective functions.
As shown in fig. 4, ψ is calculated using the least-squares method share In the time, the function image of the single-objective optimization function min { obedience (ψ), autonomy (ψ), stability (ψ) } in the value interval satisfying the constraint condition is shown as a black bold line in fig. 4, where the angle corresponding to the maximum value of the function values is used as the optimal solution of the multi-objective optimization problem, that is, the control command ψ of the ARV heading angle generated based on the sharing control method of multi-objective optimization in this embodiment share . Psi obtained using the least squares method share The objective function with the smallest function value can always obtain the maximum value in the constraint condition, the objective with the lowest objective function value is ensured to obtain the best result, and the influence of the objective function on the optimization result is conveniently changed by adjusting the shape coefficient of the objective function.
The working procedure of the method of the present invention is shown in FIG. 5, and is divided into the following steps.
(1) Designing proper objective functions and constraint conditions according to task demands, performance demands and the like, modeling the sharing control of the ARV as a multi-objective optimization problem, and selecting a proper optimization method;
(2) Determining each objective function according to the information such as the motion state of the ARV, the operation of an operator, the environmental condition and the like in the current control period;
(3) Determining constraint conditions of the shared control command according to the environmental information, the requirement on ARV motion and other information in the current control period;
(4) Calculating a sharing control command of the ARV generated by a sharing control method based on multi-objective optimization according to a pre-selected optimization method;
(5) Transmitting the sharing control command to an executing mechanism of the ARV;
(6) Judging whether the process is finished, if so, exiting, otherwise, re-executing the step (2).
In addition to the above embodiments, the method of the present invention may also have other embodiments such as the movement of the end of the ARV mechanical arm, and all the technical solutions formed by adopting equivalent substitution or equivalent transformation fall within the scope of protection required by the method of the present invention.
Claims (7)
1. The sharing control method of the autonomous remote control underwater robot based on the multi-objective optimization is characterized by comprising the following steps of:
1) The current position P of the ARV detected by the operator on the basis of the sensor ARV Heading angle psi ARV And the distribution position P of the obstacle in the local environment obs Operating the operating lever, and calculating an ARV heading angle control command psi of manual operation according to the input of the operating lever h And control command v of forward speed h ；
2) According to P obs And P ARV Calculating an autonomous ARV heading angle control command ψ apf ；
3) Constructing a shared objective function, constraint conditions and an optimization solving method thereof, and according to the current position P of the ARV ARV Heading angle psi ARV And the distribution position P of the obstacle in the local environment obs And a calculated manually operated ARV heading angle control command ψ h Autonomous ARV heading angle control command ψ apf Outputting a heading angle sharing control command psi share ；
The shared objective function comprises a compliance function obedience (psi), an autonomy (psi) and a stability function stability (psi) and is used for evaluating different values of a heading angle control command variable psi in the following forms:
aim(ψ)＝exp(-χ|ψ-ψ aim |)
in the formula, aim represents an objective function, ψ aim For the optimization objective of the objective function, exp () is expressed asNatural constants as an exponential function of the base; x is more than or equal to 0; the value ranges of the objective functions of this form are all in the interval (0, 1]An inner part; the magnitude of the function value of the objective function and the heading angle control command variable psi meet the corresponding objective psi aim Is positive in degree of correlation, at symmetry axis ψ=ψ aim The maximum value 1 is obtained and is decreased to two sides, the decreasing speed of the function value can be changed by adjusting the shape coefficient χ, the greater the χ is, the faster the decreasing speed is, and particularly, when χ=0, the function value is constant to be 1;
constraint condition I of the shared objective function safe According to the security evaluation function, the method comprises the following steps of:
firstly, determining a set I 'of value intervals of a motion direction meeting the safety requirement' safe ＝[I′ 1 ，I′ 2 ，...，I′ j ]J is the number of the value intervals, so that ARV is in I' safe A safe distance from the obstacle in the movement direction in the interval of (a); using security (ψ) < λ (d) max -d share ) Calculating to obtain I' safe ；
Wherein the security assessment function is defined as a piecewise function, and the function value on each piece of domain is determined according to the following formula:
where security (ψ) represents a security evaluation function; μ is a scaling factor changing the magnitude of the function value, set to a constant greater than 0; d, d max A distance threshold value that causes a change in the security (ψ) function value for the obstacle to start, andd is the minimum distance from ARV to the obstacle grid within a single definition domain, when d > d max Time d=d max ；
Secondly, setting a safety interval I for guaranteeing ARV safety Δ Reduce I' safe To obtain a set I of feasible solution value intervals safe ＝[I 1 ，I 2 ，...，I m ]M is the number of the value intervals, and the ARV is ensured to be in I safe The distance between the barrier on two sides in the movement direction in the interval of (2) is relatively safe;
4) ARV control command v according to forward speed h Control command psi for sharing heading angle share And controlling the ARV to avoid the obstacle and executing the task.
2. The multi-objective optimization-based autonomous remote control underwater robot sharing control method according to claim 1, wherein the ARV heading angle control command ψ is calculated manually h And control command v of forward speed h The formula of (2) is as follows:
wherein t is the time; j (J) x And J y For operator control input on the x and y axes of the lever, where J x Control of ARV heading angle, J y Control the forward speed of ARV, and J x ∈[-1，1]、J y ∈[-1，1]；ψ ARV The current heading angle of the ARV; k (k) ψ Gain factors for adjusting the heading angle increment; v mnax Is the maximum forward speed of the ARV; j (J) xd Not less than 0 and J yd And ≡0 is a dead zone threshold set near the x-and y-axis 0 positions of the lever, respectively.
3. An autonomous remote underwater robot sharing control method based on multi-objective optimization according to claim 1, characterized in that said method is based on P obs And P ARV Calculating ARV heading angle control command psi of autonomous control system apf Comprising the following steps:
(1) if present in the environmentAn obstacle, an ARV heading angle control command ψ of an autonomous control system is calculated using apf ：
ψ apf (t)＝atan(y v -y o ，x v -x o )
Wherein P is ARV ＝[x v ，y v ]For ARV current position coordinates, [ x ] o ，y o ]Is the position of the obstacle;
(2) if there are multiple obstacles in the environment, then:
calculating the magnitude r of the effect of the ith obstacle on ARV motion using the following formula i ：
Wherein i is the number of the obstacle, d obs_i For Euclidean distance, d, between ARV and the ith obstacle snare 、d safe 、k o D is a design parameter share > 0 is the maximum distance that the obstacle has to affect on ARV navigation safety, d safe > 0 is the distance threshold value, d, for ensuring safety between ARV and obstacle share 、d safe K is determined according to the radius of motion of ARV o (1≥k o > 0) is a scaling factor for adjusting the degree to which autonomous obstacle avoidance behavior assists other behavior in obstacle avoidance;
by r i Is the size of the module, ψ apf_i Determining for direction a vector V of the effect of the ith obstacle on ARV obs_i The vector sum of all the obstacle's influence on the ARV is obtained by vector addition using the following equation, using the direction of the vector sum as the control command ψ outputted from the autonomous control module apf ：
Where n is the number of obstacles in the environment that affect the safety of ARV navigation.
4. The multi-objective optimization-based autonomous remote control underwater robot sharing control method according to claim 1, wherein a compliance function obedience (ψ) is used for evaluating ψ compliance aim The degree of (3):
obedience(ψ)＝exp(-α|ψ-ψ h |)
alpha is the shape factor of the compliance function, the value of which is determined by:
wherein alpha is max A constant greater than 0; d, d mmin Is the minimum distance of the ARV to the obstacle; the formula is according to d min The value of alpha is adjusted to change the extent to which the operator affects the ARV motion.
5. The multi-objective optimization-based autonomous remote underwater robot sharing control method according to claim 1, wherein the autonomous degree function autonomy (ψ) is used for evaluating ψ and ψ apf Is the difference between (a):
autonomy(ψ)＝exp(-γ|ψ-ψ apf |)
gamma is the shape factor of the autonomous function, the value of which is determined by:
wherein, gamma max For constants greater than 0, the formula is according to d mmin Adjusting the value of gamma to make phi apf At d only mmin ＜d share The ARV is moved away from the obstacle.
6. The multi-objective optimization-based autonomous remote underwater robot sharing control method according to claim 1, wherein the stability function stability (ψ) is used for evaluating the relative value ψ ARV The degree of variation of (2):
stability(ψ)＝exp(-β|ψ-ψ ARV |)
where β is the shape factor of the stability function and is a normal number.
7. The multi-objective optimization-based autonomous remote control underwater robot sharing control method is characterized in that the optimization solving method of the sharing objective function adopts a minimum maximum method, and a heading angle control command psi generated by the sharing control method can be obtained through solving share The following formula is shown:
where min { } represents that each segment in the defined domain takes the minimum of all objective functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910401020.2A CN112035992B (en) | 2019-05-14 | 2019-05-14 | Multi-objective optimization-based autonomous remote control underwater robot sharing control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910401020.2A CN112035992B (en) | 2019-05-14 | 2019-05-14 | Multi-objective optimization-based autonomous remote control underwater robot sharing control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112035992A CN112035992A (en) | 2020-12-04 |
| CN112035992B true CN112035992B (en) | 2024-01-09 |
Family
ID=73575738
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910401020.2A Active CN112035992B (en) | 2019-05-14 | 2019-05-14 | Multi-objective optimization-based autonomous remote control underwater robot sharing control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112035992B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114779801B (en) * | 2021-01-22 | 2024-05-31 | 中国科学院沈阳自动化研究所 | Autonomous remote control underwater robot path planning method for target detection |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0896267A2 (en) * | 1997-08-04 | 1999-02-10 | Fuji Jukogyo Kabushiki Kaisha | Position recognizing system of autonomous running vehicle |
| JP2009264965A (en) * | 2008-04-25 | 2009-11-12 | Mitsubishi Heavy Ind Ltd | Underwater sailing body and obstacle detection apparatus |
| KR20120129002A (en) * | 2011-05-18 | 2012-11-28 | 부산대학교 산학협력단 | Underwater robot and Method for controlling the same |
| CN103116279A (en) * | 2013-01-16 | 2013-05-22 | 大连理工大学 | Vague discrete event shared control method of brain-controlled robotic system |
| KR20160099965A (en) * | 2015-02-13 | 2016-08-23 | 삼성중공업 주식회사 | Apparatus and method for recognizing position of underwater vehicle |
| CN105955268A (en) * | 2016-05-12 | 2016-09-21 | 哈尔滨工程大学 | Local obstacle avoidance considering UUV moving object sliding mode tracking control method |
| CN106959698A (en) * | 2017-05-24 | 2017-07-18 | 大连海事大学 | A kind of path trace avoidance method of guidance |
| JP2017151499A (en) * | 2016-02-22 | 2017-08-31 | 株式会社Ihi | Obstacle avoidance method and device |
| CN108829134A (en) * | 2018-07-03 | 2018-11-16 | 中国船舶重工集团公司第七〇九研究所 | A kind of real-time automatic obstacle avoiding method of deepwater robot |
| CN109540151A (en) * | 2018-03-25 | 2019-03-29 | 哈尔滨工程大学 | A kind of AUV three-dimensional path planning method based on intensified learning |
-
2019
- 2019-05-14 CN CN201910401020.2A patent/CN112035992B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0896267A2 (en) * | 1997-08-04 | 1999-02-10 | Fuji Jukogyo Kabushiki Kaisha | Position recognizing system of autonomous running vehicle |
| JP2009264965A (en) * | 2008-04-25 | 2009-11-12 | Mitsubishi Heavy Ind Ltd | Underwater sailing body and obstacle detection apparatus |
| KR20120129002A (en) * | 2011-05-18 | 2012-11-28 | 부산대학교 산학협력단 | Underwater robot and Method for controlling the same |
| CN103116279A (en) * | 2013-01-16 | 2013-05-22 | 大连理工大学 | Vague discrete event shared control method of brain-controlled robotic system |
| KR20160099965A (en) * | 2015-02-13 | 2016-08-23 | 삼성중공업 주식회사 | Apparatus and method for recognizing position of underwater vehicle |
| JP2017151499A (en) * | 2016-02-22 | 2017-08-31 | 株式会社Ihi | Obstacle avoidance method and device |
| CN105955268A (en) * | 2016-05-12 | 2016-09-21 | 哈尔滨工程大学 | Local obstacle avoidance considering UUV moving object sliding mode tracking control method |
| CN106959698A (en) * | 2017-05-24 | 2017-07-18 | 大连海事大学 | A kind of path trace avoidance method of guidance |
| CN109540151A (en) * | 2018-03-25 | 2019-03-29 | 哈尔滨工程大学 | A kind of AUV three-dimensional path planning method based on intensified learning |
| CN108829134A (en) * | 2018-07-03 | 2018-11-16 | 中国船舶重工集团公司第七〇九研究所 | A kind of real-time automatic obstacle avoiding method of deepwater robot |
Non-Patent Citations (2)
| Title |
|---|
| Dynamic Shared Control for Human-Wheelchair Cooperation;Qinan Li et al;2011 IEEE International Conference on Robotics and Automation;第4278-4283页 * |
| 自主/遥控水下机器人（ARV）共享控制研究;刘重洋;中国优秀硕士学位论文全文数据库 信息科技辑(第12期);I140-520 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112035992A (en) | 2020-12-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Research on fuzzy control of path tracking for underwater vehicle based on genetic algorithm optimization | |
| CN105955268B (en) | A kind of UUV moving-target sliding mode tracking control methods considering Local obstacle avoidance | |
| CN109784201B (en) | AUV dynamic obstacle avoidance method based on four-dimensional risk assessment | |
| CN112527019B (en) | Heterogeneous unmanned system cooperative formation control system and method suitable for severe sea conditions | |
| CN108061897A (en) | A kind of submerged structure environment line feature extraction method based on Forward-Looking Sonar | |
| Wei et al. | 3D semantic map-based shared control for smart wheelchair | |
| Haugaløkken et al. | Monocular vision-based gripping of objects | |
| CN112035992B (en) | Multi-objective optimization-based autonomous remote control underwater robot sharing control method | |
| bin Mansor | Motion control algorithm for path following and trajectory tracking for unmanned surface vehicle: A review paper | |
| Rodrigues et al. | Online range-based slam using b-spline surfaces | |
| Zhang et al. | Multi-UUV path planning based on improved artificial potential field method | |
| Jiang et al. | A method for underwater human–robot interaction based on gestures tracking with fuzzy control | |
| Hamad et al. | Path Planning of Mobile Robot Based on Modification of Vector Field Histogram using Neuro-Fuzzy Algorithm. | |
| Kurz | ALEF: An autonomous vehicle which learns basic skills and constructs maps for navigation | |
| Xu et al. | Trajectory tracking for underwater rescue salvage based on backstepping control | |
| CN116430856A (en) | Ship track control method based on variable forward looking distance LOS algorithm | |
| CN114200966B (en) | Unmanned aircraft target orientation equidistant tracking method based on perception information | |
| Zhou et al. | Visual servo control of underwater vehicles based on image moments | |
| Chu et al. | Design of a dead reckoning based motion control system for small autonomous underwater vehicle | |
| Aubard et al. | LSTS Toolchain Framework for Deep Learning Implementation into Autonomous Underwater Vehicle | |
| CN108170976B (en) | Safety analysis method in AUV dynamic recovery process of underwater submarine | |
| Wei et al. | Semantic mapping for smart wheelchairs using rgb-d camera | |
| WO2021045679A1 (en) | Method and system for object localization, and controlling movement of an autonomous underwater vehicle for object intervention | |
| Nejad et al. | Real time localization, path planning and motion control for autonomous parking in cluttered environment with narrow passages | |
| Sanchez et al. | Autonomous navigation with deadlock detection and avoidance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |