CN113361087B - Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle - Google Patents
Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle Download PDFInfo
- Publication number
- CN113361087B CN113361087B CN202110603204.4A CN202110603204A CN113361087B CN 113361087 B CN113361087 B CN 113361087B CN 202110603204 A CN202110603204 A CN 202110603204A CN 113361087 B CN113361087 B CN 113361087B
- Authority
- CN
- China
- Prior art keywords
- layout
- sensor
- underwater vehicle
- sensors
- optimization
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a method and a system for optimizing the position layout of a lateral line detection sensor of an underwater vehicle, wherein a real pressure measurement value is expressed according to an obtained theoretical pressure data set and a prediction error and a probability density function of a likelihood function is represented according to a physical model of the underwater vehicle and the flow field environment of the physical model of the underwater vehicle; providing a high-efficiency integrated sequential heuristic optimization algorithm to complete the calculation of the optimized array layout of the sensors; the optimization simulation analysis is carried out on the array distribution on the basis of no experiment, and a theoretical basis is laid for the establishment of a lateral line detection experiment carrier and an experiment platform.
Description
Technical Field
The invention relates to the field of sensor array optimization, in particular to a method and a system for optimizing the position layout of a lateral line detection sensor of an underwater vehicle.
Background
The survival of fishes in water depends on the collection and perception of underwater environment information in the nature, a visual system and a lateral line perception system are the two most main sensing organs of the fishes, wherein the lateral line system provides inspiration for the design of an artificial lateral line system of an underwater vehicle, in the conventional research on the layout of a bionic lateral line sensor array, the head and the side of an underwater robot are mainly distributed at equal intervals, the underwater robot is set according to experience and then experimental data acquisition is carried out, a mathematical model and theoretical basis are lacked to carry out the optimization guidance on the distribution of the positions of the sensors, the accuracy of judgment and prediction of the positions of dipole vibration and moving target objects and the positioning and detection efficiency of the target objects are influenced, so the advantages and the disadvantages of the distribution modes of the sensors are required to be judged, and the optimized layout scheme of the sensors is obtained through the mathematical theory guidance.
Disclosure of Invention
The invention aims to provide a method and a system for optimizing the position layout of a side line detection sensor of an underwater vehicle, which are used for overcoming the defects in the prior art.
In order to achieve the purpose, the invention adopts the following technical scheme:
a method for optimizing the position layout of a lateral line detection sensor of an underwater vehicle comprises the following steps:
s1, establishing an underwater vehicle physical model and a flow field environment of the underwater vehicle physical model;
s2, setting initial layout observation points of the sensors based on the physical model of the underwater vehicle and the flow field environment of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
s3, representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error and representing a probability density function of a likelihood function;
s4: carrying out relative entropy calculation on the likelihood function in the S3, and predicting the relative entropies of the prior probability distribution and the posterior probability distribution of the motion position of the target object and the expected values of all the relative entropies to serve as target functions;
s5: converting the unknown quantity in the target function by using Bayes theorem to obtain a converted target function;
s6: comparing the size of the converted target function to obtain a first optimization point position, and sequentially calculating by a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point;
s7: and optimizing the distribution number of the sensor arrays by calculating the spearman grade correlation coefficient of each arrangement to obtain the optimal number of the sensors, thereby completing the position layout optimization of the lateral line detection sensors of the underwater vehicle.
Further, the flow field environment includes a target position, a motion parameter, and a flow field area parameter.
Further, simulation of the underwater vehicle physical model under the flow field environment is performed through establishing a flow field model, determining a calculation domain, dividing a calculation grid, setting solving parameters, iterative calculation and post-processing, and the oscillation of dipoles rs at each position of different pressure observation points si is obtainedThe pressure monitoring curve of the generated water flow environment is used for obtaining the pressure value of each sensor initial layout point S to each target object position rs as a theoretical calculation value F (r) of the array optimization algorithm s (ii) a s), generating a theoretical pressure data set.
Further, the prediction error ε(s) is the true measurement y and the theoretical measurement F (r) s (ii) a s), i.e.:
y＝F(r s ；s)+ε(s) (3)。
further, the optimized sensor distribution mode is determined by calculating the utility function value under each layout:
p (r) is the prior probability, p (y | r, s) is the likelihood function, p (y | s) is the probability distribution, y is the pressure sensor measurement, and p (y | r, s) is the probability distribution of the pressure sensor measurement y.
Further, the objective function transformed by using bayes theorem is as follows:
and integrating the interference source position variable r to obtain p (y | s).
Further, the step of sequentially calculating the array layout optimization position coordinate points through a sequential heuristic array optimization algorithm specifically comprises the following steps:
step 6.1: calculating the size of a target function of an initial layout observation point of each sensor, determining a position point corresponding to the maximum value of the target function, and using the position point as a first sensor position point for array layout optimization;
step 6.2: respectively combining the first sensor position point and the rest initial sensor position points, and respectively calculating a target function of the combined array coordinates for predicting the motion position of the target object;
step 6.3: comparing the size of the target function, and selecting the position point combination of the maximum value of the target function as the selection result of the position points of the two sensors before the array optimization;
step 6.4: and 6.1-6.3 are repeated to complete the selection of the position points of the plurality of sensors.
Further, specifically, a plurality of sensors are arranged on one side of the distribution model, the number of the sensors is gradually increased, meanwhile, the information entropy change is guaranteed to be maximum, and the optimal overall layout of the sensors is determined by selecting one sensor distribution position point at a time.
Further, a sensor position point with the largest information entropy reduction amplitude is selected as the optimal position S of the first sensor under the condition of having the initial layout of the sensors 1 Passing the optimum position S of the first sensor on the basis of the optimum position of the first sensor 1 The entropy change of the information obtained by combining the position of the second sensor is calculated, and the sensor combination with the maximum change, namely the maximum objective function, is the optimal position S of the first two sensors 1 And S 2 Obtaining the optimal distribution combination (S) of the sensors 1 ,S 2 )。
An underwater vehicle siding detection sensor position layout optimization system comprising:
the model establishing module is used for establishing an underwater vehicle physical model and setting a flow field environment of the underwater vehicle physical model;
the theoretical pressure generation module is used for setting initial layout observation points of the sensors according to the physical model of the underwater vehicle and the flow field environment of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
the target function optimization module is used for representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error, representing a probability density function of the likelihood function, calculating relative entropy of the likelihood function, taking the relative entropy of the prior probability distribution and the posterior probability distribution of the predicted target object motion position and the expected values of all the relative entropies as the target function, converting the unknown quantity in the target function to obtain a converted target function, comparing the size of the converted target function to obtain a first optimization point, sequentially calculating through a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point, optimizing the distribution number of the sensor arrays by calculating the Spanish scale correlation coefficient of each arrangement to determine the optimal number of the sensors, and accordingly completing the optimization of the position layout of the underwater vehicle lateral line detection sensors.
Compared with the prior art, the invention has the following beneficial technical effects:
the invention relates to a position layout optimization method of a lateral line detection sensor of an underwater vehicle, which is characterized in that an initial layout observation point of the sensor is set according to a physical model of the underwater vehicle and a flow field environment of the physical model of the underwater vehicle, a theoretical pressure data set is generated after simulation, then a real pressure measurement value is expressed according to the theoretical pressure data set and a prediction error, a probability density function of a likelihood function is represented, and a theoretical basis is provided for the optimized distribution of a lateral line sensor array through a Bayesian probability formula and a sequential heuristic optimization algorithm in combination with a correlation coefficient representation method; the invention adopts a Bayesian probability theory model to well convert the target detection problem into the probability problem, namely, the sensor array optimization layout problem is converted into certain array distribution, so that the most accurate estimation and judgment can be made on the motion position of the target object; providing a high-efficiency integrated sequential heuristic optimization algorithm to complete the calculation of the optimized array layout of the sensors; and optimizing simulation analysis is carried out on the array distribution on the basis of no experiment, and a theoretical basis is laid for the establishment of a lateral line detection experiment carrier and an experiment platform.
Furthermore, by adopting the sequential heuristic algorithm, higher accuracy level can be achieved by less calculation amount.
Furthermore, the likelihood function is adopted to calculate the relative entropy, so that the position of the vibration interference source can be estimated most accurately, and the optimization accuracy is improved.
Drawings
Figure 1 is a diagram of physical model dimensions of an underwater vehicle in an embodiment of the invention.
Fig. 2 is a schematic diagram of the dimensional and positional parameters of the flow field region in an embodiment of the present invention.
FIG. 3 is a graph of initial point locations of a sensor array in an embodiment of the invention.
Fig. 4 is a flowchart of a sequential heuristic layout optimization algorithm in an embodiment of the present invention.
Fig. 5 is a flow chart of layout optimization of a sensor array in an embodiment of the invention.
FIG. 6 is a flowchart of a probabilistic validation optimization layout in an embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
as shown in fig. 5, a method for optimizing the position layout of side line detection sensors of an underwater vehicle comprises the following steps:
s1, establishing a physical model of an underwater vehicle, and determining the position, the motion parameters and the flow field area parameters of a target object;
modeling and drawing a physical model of the standard underwater vehicle according to the size of the underwater vehicle to obtain the physical model of the underwater vehicle, determining a flow field environment, setting a sensor initial layout position point s on the boundary of the physical model, wherein the flow field environment comprises a target object position, a motion parameter and a flow field area parameter; drawing a target object motion area, and determining flow field area parameters of a hydrodynamic simulation analysis flow field based on the design size of an underwater vehicle physical model, the motion form and range of an experimental target object and the size specification parameters of an experimental scene, wherein the flow field area parameters comprise the size of a square area and the position size of the central point of the front end surface of an underwater vehicle in the flow field area; and determining the motion parameters of the target object, wherein the motion parameters of the target object comprise a motion form, position coordinates, an activity area and the number of targets.
S2, setting initial layout observation points of the sensors based on the physical model of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
simulating a physical model of an underwater vehicle in the flow field environment in FLUENT fluid hydrodynamic simulation software, and establishing a flow field model, determining a calculation domain, dividing a calculation grid, setting a solving parameter and overlappingPerforming substitute calculation and post-processing to complete simulation of the physical model of the underwater vehicle in the flow field environment, obtaining a pressure monitoring curve of the water flow environment generated by the oscillation of dipoles rs at each position by different pressure observation points si, and obtaining a pressure value of each sensor initial layout point S to each target position rs as a theoretical calculation value F (r) of an array optimization algorithm s (ii) a s), generating a theoretical pressure data set.
S3, representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error and representing a probability density function of a likelihood function;
on the basis of the step S1 and the step S2, the Bayesian probability model is corresponded according to the working characteristics of the sensor lateral line array for detecting the target object in the water, and the target detection work is corresponded to the probability events one by one. Bayesian theorem is used to describe the relationship between two conditional probabilities, such as P (a | B) and P (B | a), which respectively represent the probability of occurrence of event a under the occurrence condition of event B and the probability of occurrence of event B under the occurrence condition of event a, and the bayesian formula can be obtained according to the theorem as follows:
for the detection work of the underwater survey line array target object, the current dipole subdirectory target object is set as r s (x s ,y s ) Predicting the motion position of the dipole target object as an event A, quantifying the uncertainty of the event A by a probability distribution rule, and updating the probability distribution rule according to pressure data information captured by a lateral line sensor array distributed on an underwater vehicle. As long as a lateral line array distributed on an underwater vehicle can detect a water pressure data signal generated by disturbance of dipole vibration at a position r to surrounding fluid, the position of a vibrating dipole can be detected, an event A can be judged according to the data, and an unknown pressure signal is judged to be caused by dipole vibration at the position r. Therefore, the problem of optimal distribution of the sensors can be converted into a mode of searching a sensor lateral line array distribution to make a motion position of a target object, namely an event AThe most accurate estimation and judgment.
The side line sensors are distributed at one side of the underwater vehicle, the position is represented by s (s belongs to R), and F (R) s (ii) a s) represents a predicted pressure value of the sensor s for the dipole oscillation flow field generated at the position r, and the predicted value is obtained by performing simulation calculation based on a Navier-Stokes equation on a theoretical pressure value of the sensor at the position s when the dipole oscillation occurs at the position r under the flow field environment. Let the prior probability of the event a be known, denoted as p (r), after obtaining the measured value y of the sensor, the posterior probability distribution p (r | y, s) can be determined, and according to the bayesian principle, the posterior probability distribution p (r | y, s) is proportional to the product of the corresponding prior probability distribution p (r) and the likelihood probability p (y | r, s), that is:
p(r|y,s)∝p(r)·p(y|r,s) (2)
where the likelihood probability function p (y | r, s) represents the probability that a pressure measurement y measured at a given sensor location layout, s, will result from the vibration of the interference source at r. Because the real pressure measurement value has a certain error with the theoretical pressure measurement value, the prediction error epsilon(s) is set as the real measurement value y and the theoretical measurement value (namely the prediction measurement value) F (r) s (ii) a s), i.e.:
y＝F(r s ；s)+ε(s) (3)
s4: carrying out relative entropy calculation on the likelihood function in the S3, and predicting the relative entropies of the prior probability distribution and the posterior probability distribution of the motion position of the target object and the expected values of all the relative entropies to serve as target functions;
since the maximum entropy principle provides a distribution criterion which best meets objective conditions when random variable statistical characteristics are selected, selecting the distribution with the maximum entropy as the distribution of the random variables is an effective processing method, and when the mean value and covariance matrix are constant, the random variables meet the normal distribution and have the maximum entropy, therefore, the prediction error epsilon (S) is set to meet the definition, and epsilon (S) obeys a multivariate Gaussian distribution N (0, sigma (S)) with the mean value of 0 and the covariance matrix sigma (S) being constant, therefore, according to the prediction error formula, the likelihood function p (y | r, S) of the real pressure measurement value is obtained to obey the mean value of F (r | r, S) s (ii) a s), preparation ofThe variance matrix has certain multivariate Gaussian distribution, and the expression is as follows:
and (3) calculating the relative entropy of the likelihood function obtained in the step (S3), wherein the problem of the optimal sensor layout is converted into whether the determined certain sensor layout S can carry out the most accurate estimation on the position of the vibration interference source at the position r, namely the measured value obtained under the layout S is the most effective for estimating the position S of the interference source, and in order to describe the information quantity in a mathematical mode, the concept of KL divergence (relative entropy) in the information theory is introduced. Relative entropy or information divergence is an asymmetry that measures the difference between two probability distributions, which can measure the distance between two random distributions, and when two random distributions are the same, their relative entropy is zero, and when the difference between two random distributions increases, their relative entropy also increases. The relative entropy can thus compare the similarity of the two probability distribution information. Now, reconsidering event a — finding the coordinates of the dipole target motion position in the target area, knowing that the prior distribution of event a is p (r) and that the posterior distribution after obtaining the pressure measurement value is p (ry, s), because the sensor distribution pattern s in the posterior distribution probability p (ry, s) is given as a condition, the greater the information amount measured by the pressure sensor, the greater the difference between the prior probability distribution p (r) and the posterior probability distribution p (ry, s), i.e. the greater the relative entropy, when the sensor layout pattern s is selected, the better. Thus, to give a utility function representing the information gain of the a priori and a posteriori distributions of event a, the relative entropy between them is defined as:
since the pressure sensor measurement value y cannot be obtained experimentally in the sensor layout optimization stage, it is determined by the error model equation (3)And determining the size of a sensor measurement value y under the given fixed interference source position parameter r and the sensor distribution mode parameter s. Since the information gain between the a-priori distribution and the a-posteriori distribution of the event a is represented by the relative entropy between the two, the optimal distribution of the sensors s is determined by maximizing the utility function of this information gain. The quality of the sensor layout can be reflected by the formula (5), which indicates that a certain sensor s is under a certain sensor layout s n The gain of information after obtaining the measurement value y changes, so in order to show the relative entropies of all sensors under the layout, it is necessary to improve the formula (5), calculate the expected values of the relative entropies of all possible pressure sensor measurement values as utility functions, determine the optimal layout of the sensors by maximizing the utility functions, determine the optimized sensor distribution by calculating the utility function values under each layout, and the improved function is defined as:
s5: converting the unknown quantity in the target function by using Bayes' theorem to obtain a converted target function;
and (4) performing unknown quantity conversion on the target function in the step (S4), except that the prior distribution p (r) is known, and both p (r | y, S) and p (y | S) are unknown parameters, so that a Bayesian probability formula is required to be used for converting the target function.
The bayesian probability formula is a theorem on the conditional probabilities of two events a and B, and its expression is:
according to Bayes' theorem, two events, namely the event A and the event B, which respectively correspond to the dipole target motion position and the pressure sensor measurement value, can be obtained by combining the problem, so that the formula (7) can be converted into:
part of functions in the formula (6) are transformed:
the objective function of equation (6) thus translates into:
equation (11) is an objective function transformed by using bayesian theorem, in which the prior probability p (r) is known, the likelihood function p (y | r, s) can be expressed mathematically by a multivariate gaussian function (see equation (4)), and only the probability distribution p (y | s) is an unknown parameter. The probability distribution represents the probability distribution of the pressure sensor measurement value y when the known sensor layout is s, the probability distribution of the pressure sensor measurement value y is known to be p (y | r, s), and p (y | s) can be obtained by integrating the interference source position variable r, and the mathematical expression is as follows:
therefore, all unknown variables in the objective function are converted into known parameters and defined by mathematical expressions as follows:
s6: comparing the size of the converted target function to obtain a first optimization point and sequentially calculating by a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point;
and (5) calculating the target function converted in the step (S5) by a sequential heuristic optimization algorithm, which comprises the following specific steps:
step 6.1: calculating the size of a target function of an initial layout observation point of each sensor, determining a position point corresponding to the maximum value of the target function, and using the position point as a first sensor position point for array layout optimization;
step 6.2: respectively combining the first sensor position point and the rest initial sensor position points, and respectively calculating a target function of the combined array coordinates for predicting the motion position of the target object;
step 6.3: comparing the size of the target function, and selecting the position point combination of the maximum value of the target function as the selection result of the position points of the two sensors before the array optimization;
step 6.4: and 6.1-6.3 are repeated to complete the selection of the position points of the plurality of sensors.
The method comprises the following specific steps: and (3) obtaining the target function formula (13), finishing theoretical derivation of the sensor optimization layout model, and importing and calculating initial distribution coordinate points of the sensor to obtain the corresponding optimization layout. The optimal sensor layout mode is obtained by maximizing the objective function in the formula (13), however, the optimal sensor layout problem is characterized in that a relatively large number of local optimal solutions exist, the respective calculation of the local optimal solutions under various sensor number combinations is unscientific and inefficient, and the use of optimization algorithms such as genetic algorithms with wide application requires specific optimization parameters, which leads to a result of over-parameters, so that a more effective algorithm needs to be found for sensor distribution optimization.
Is provided with a fixed number N 0 The sensor distribution combination is known, the initial distribution position points of the sensors before optimization are known, the sequential heuristic sensor layout algorithm is a more effective and more integrated sensor distribution optimization method, and compared with a genetic algorithm, the sequential heuristic algorithm can achieve a higher precision level through less calculation amount. By arranging a plurality of sensors on one side of the distribution model, the number of the sensors is gradually increased while the information entropy change is ensured to be maximum, and one sensor is selected at a timeThe location points are distributed to determine an optimized overall layout of the sensors. Specifically, first, a sensor position point with the largest information entropy reduction, that is, a position point with the largest objective function value is selected as the optimal position S of the first sensor under the condition of having the initial sensor layout 1 Passing the optimum position S of the first sensor on the basis of the optimum position of the first sensor 1 The entropy change of the information obtained by combining the position of the second sensor is calculated, and the sensor combination with the maximum change, namely the maximum objective function, is the optimal position S of the first two sensors 1 And S 2 Obtaining the optimal distribution combination (S) of the sensors 1 ,S 2 ) The method is a sequential heuristic sensor optimization algorithm. Continuing the calculation in the same way, when obtaining the distribution combination of the optimal position distribution points of i-1 sensors and wanting to obtain the ith optimal distribution sensor position point, selecting the ith point with the minimum information entropy change quantity obtained by the i sensor position combinations as the distribution position point of the optimal sensor, namely obtaining the optimal distribution combination of the i sensors (S) 1 ，S 2 ，……，S i ). For a maximum of N 0 And (4) continuously and repeatedly circulating the processes by the sensors. For 1 to N 0 The sequential sensor placement algorithm will give the optimal sensor configuration only if the optimal sensor positions for the i sensors are a subset of the optimal sensor positions for the i +1 sensors. And continuously repeating the steps until the optimal distribution number i reaches a preset value or the number of the sensors is increased, and finishing the optimal calculation of the sensor distribution to obtain the optimal layout of the sensors, wherein the reduction of the information entropy is small, namely the difference of the objective function values calculated by the combination of the two sensors before and after the increase of the number of the sensors is not large. The flow chart of the sequential heuristic sensor array layout optimization algorithm is shown in FIG. 4, and the flow chart of the overall layout optimization is shown in FIG. 5.
S7: and optimizing the distribution number of the sensor arrays to obtain the optimal number of the sensors by calculating the spearman grade correlation coefficient of each arrangement, thereby completing the position layout optimization of the lateral line detection sensors of the underwater vehicle.
After the optimization of the array position is finished in step S6, the distribution S of each sensor array is calculated 1 ,S 2 ,S 3 ,……，S i Next, for a certain interference source location r s Posterior probability distribution p (r) obtained by position detection s Y, S) as a certain distribution S i A posterior probability density function of time, and calculating S i And S i+1 (i∈[1，n-1]) The correlation coefficient R of the adjacent arrays is calculated, when S is i ≥S n When R is infinitely close to 1 and is kept stable, it is considered that the increase of the number of sensor points is unnecessary, and therefore, the optimum number of sensors is determined as S n 。
After the number of the sensors in the step 7 is determined, the posterior probability function value p (r) under each array distribution is obtained through calculation s Y, S) by a certain distribution S i And (3) performing probability representation of each possible point position of the lower detection area, and calculating coordinates of vibration position points of the predicted target object to achieve the purpose of optimizing distribution inspection of the sensor array.
In one embodiment of the invention, a terminal device is provided that includes a processor and a memory, the memory storing a computer program comprising program instructions, the processor being configured to execute the program instructions stored by the computer storage medium. The processor is a Central Processing Unit (CPU), or other general purpose processor, digital Signal Processor (DSP), application Specific Integrated Circuit (ASIC), ready-made programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc., which is a computing core and a control core of the terminal, and is adapted to implement one or more instructions, and in particular, to load and execute one or more instructions to implement a corresponding method flow or a corresponding function; the processor provided by the embodiment of the invention can be used for realizing the operation of the optimization method of the position layout of the side line detection sensors of the underwater vehicle.
The embodiment is as follows: a position layout optimization system for a lateral line detection sensor of an underwater vehicle comprises a model establishing module, a side line detection module and a side line detection module, wherein the model establishing module is used for establishing a physical model of the underwater vehicle and setting a flow field environment of the physical model of the underwater vehicle;
the theoretical pressure generation module is used for setting initial layout observation points of the sensors according to the physical model of the underwater vehicle and the flow field environment of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
the target function optimization module is used for representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error, representing a probability density function of the likelihood function, calculating relative entropy of the likelihood function, taking the relative entropy of the prior probability distribution and the posterior probability distribution of the predicted target object motion position and the expected values of all the relative entropies as the target function, converting the unknown quantity in the target function to obtain a converted target function, comparing the size of the converted target function to obtain a first optimization point, sequentially calculating through a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point, optimizing the distribution number of the sensor arrays by calculating the Spanish scale correlation coefficient of each arrangement to determine the optimal number of the sensors, and accordingly completing the optimization of the position layout of the underwater vehicle lateral line detection sensors.
In still another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a Memory device in the terminal device and is used for storing programs and data. The computer-readable storage medium includes a built-in storage medium in the terminal device, provides a storage space, stores an operating system of the terminal, and may also include an extended storage medium supported by the terminal device. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer-readable storage medium may be a high-speed RAM memory, or may be a Non-volatile memory (Non-volatile memory), such as at least one disk memory. One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to perform the corresponding steps of the method for optimizing a layout of underwater vehicle side line detection sensor positions in the above embodiments.
Modeling and drawing a physical model of the underwater vehicle according to the size of the underwater vehicle, wherein the physical model of the underwater vehicle is shown in figure 1, and setting an initial sensor layout position point s on the boundary of the physical model; a target object motion area is drawn up, the size of a square area of a hydrodynamic simulation analysis flow field is regulated to be 1.5m multiplied by 1.5m, the central point of the front end face of the underwater vehicle is located at a certain position of the flow field area (as shown in fig. 2), and the specific flow field area parameter is regulated as shown in fig. 2. After the structural size of the underwater vehicle is determined, initial sensor layout position points s, namely pressure value observation points in hydrodynamic flow field simulation, need to be determined, boundary surfaces of the underwater vehicle are divided into three parts, namely side surfaces, top surfaces and bottom surfaces, wherein the side surfaces are provided with one observation point every 2mm, 45 observation points are totally arranged, x coordinates of the upper surface and the lower surface are provided with one observation point every 2mm, 45 observation points are respectively arranged, and the observation points at the connecting points are added, so that the total number of 137 initial observation point positions is shown in fig. 3. The method comprises the steps that a vibrating dipole is adopted as a target object for detection, the position area of the dipole target object is 1m multiplied by 1m and is located at 0.1m on the left side of the underwater vehicle, each interval of 0.1m between the horizontal coordinate and the vertical coordinate is a dipole sub-target vibration position point r, the vibration frequency is 5Hz, and the specific target object moving area is shown as the dotted line area in figure 2.
And in FLUENT fluid hydrodynamic simulation software, performing analog simulation on the flow field environment to obtain a pressure monitoring curve of the water flow environment generated by the oscillation of dipoles rs at each position by different pressure observation points Si, obtaining a simulation theoretical pressure value of each sensor initial layout point Si on each target position rs, and performing layout optimization calculation by taking the fundamental frequency amplitude of each theoretical value as a characteristic value.
Through the modeling and simulation analysis, the initial distribution S of the sensor array is determined 1 -S 137 Expressing the real pressure measurement value by using the simulation theoretical value and the prediction error and characterizing the probability density function of the likelihood functionCalculating relative entropies of prior probability distribution and posterior probability distribution of predicted target object motion position and expected values of all relative entropies as target function, converting unknown quantity in the target function by using Bayes' theorem, comparing sizes of the target function to obtain first optimized point position, and comparing target function U 1 (s) size to give U max1 (s)＝U 1 (S 121 ) =1.7393, and sequentially calculating by a sequential heuristic array optimization algorithm to obtain array layout optimization position points, in this example, the loop iteration calculation is stopped after 10 times, that is, the number of the obtained sensor optimization arrays is 10, and the calculation result is [121,11,113,133,129,125,41,31,115,1 ]]. The spearman correlation coefficients for each distribution were calculated to be [0.5369,0.9420,0.9868,0.9877,0.9872,0.9537,0.9959,0.9746,0.9927, respectively]When the 9 th point and the 10 th point are taken, the correlation coefficient is very close to 1 and keeps the same trend, so that the first 8 sensor points, namely [121,11,113,133,129,125,41,31 ] are finally selected]As a result of the array optimization arrangement.
And finally, detecting accuracy of the sensor array by calculating and representing posterior probability distribution. A flowchart of the probability distribution testing method is shown in fig. 6.
In summary, in the method for optimizing the position layout of the sensor of the underwater vehicle side line detection array, in the embodiment, the fluid simulation software is adopted to calculate the theoretical pressure value, the theoretical value is more accurately calculated, meanwhile, the error between the real value and the theoretical value is represented by probability distribution in a mathematical expression mode, a theoretical basis is provided for the optimization of the sensor array, the adaptability adjustment can be carried out according to the detection environment and the structure of the underwater vehicle, and a basis for acquiring pressure data and accurately positioning the pressure data is provided for the side line detection target object of the underwater vehicle.
The above-mentioned contents are only for explaining the technical idea of the invention of the present application, and can not be used as the basis for limiting the protection scope of the invention, and any modifications and substitutions made on the technical solution according to the design concept and technical features proposed by the present invention are within the protection scope of the claims of the present invention.
Claims (10)
1. A method for optimizing the position layout of a lateral line detection sensor of an underwater vehicle is characterized by comprising the following steps:
s1, establishing an underwater vehicle physical model and a flow field environment of the underwater vehicle physical model;
s2, setting initial layout observation points of the sensors based on the physical model of the underwater vehicle and the flow field environment of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
s3, representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error and representing a probability density function of the likelihood function;
s4: performing relative entropy calculation on the likelihood function in the S3, and predicting the relative entropy of the prior probability distribution and the posterior probability distribution of the motion position of the target object and the expected values of all the relative entropies to serve as target functions;
s5: converting the unknown quantity in the target function by using Bayes' theorem to obtain a converted target function;
s6: comparing the size of the converted target function to obtain a first optimization point and sequentially calculating by a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point;
s7: and optimizing the distribution number of the sensor arrays to obtain the optimal number of the sensors by calculating the spearman grade correlation coefficient of each arrangement, thereby completing the position layout optimization of the lateral line detection sensors of the underwater vehicle.
2. The method for optimizing the position layout of the underwater vehicle side line detection sensor according to claim 1, wherein the flow field environment comprises a target position, a motion parameter and a flow field area parameter.
3. The method for optimizing the position layout of the underwater vehicle side line detection sensors according to claim 1, wherein the method is implemented by establishing a flow field model, determining a calculation domain, dividing a calculation grid and settingSolving parameters, iterative calculation and post-processing are set, simulation is carried out on a physical model of the underwater vehicle under the flow field environment, a pressure monitoring curve of the water flow environment generated by oscillation of dipoles rs at each position of different pressure observation points si is obtained, and the pressure value of each initial layout point S of each sensor to each target position rs is obtained and used as a theoretical calculation value F (r) of an array optimization algorithm s (ii) a s), generating a theoretical pressure data set.
4. The method for optimizing the position layout of the underwater vehicle side line detection sensors as claimed in claim 1, wherein the prediction error e(s) is a real measured value y and a theoretical measured value F (r) s (ii) a s), i.e.:
y＝F(r s ；s)+ε(s) (3)。
5. the method for optimizing the position layout of the underwater vehicle side line detection sensors according to claim 1, wherein the optimized distribution mode of the sensors is determined by calculating the utility function value under each layout:
p (r) is the prior probability, p (y | r, s) is the likelihood function, p (y | s) is the probability distribution, y is the pressure sensor measurement, and p (y | r, s) is the probability distribution of the pressure sensor measurement y.
6. The method for optimizing the position layout of the underwater vehicle side line detection sensors according to claim 5, wherein the objective function transformed by Bayesian theorem is as follows:
and integrating the interference source position variable r to obtain p (y | s).
7. The method for optimizing the position layout of the underwater vehicle side line detection sensor according to claim 6, wherein the step of sequentially calculating the array layout optimized position coordinate points through a sequential heuristic array optimization algorithm specifically comprises the following steps:
step 6.1: calculating the size of a target function of an initial layout observation point of each sensor, determining a position point corresponding to the maximum value of the target function, and using the position point as a first sensor position point for array layout optimization;
step 6.2: respectively combining the first sensor position point and the rest initial sensor position points, and respectively calculating a target function of the combined array coordinates for predicting the motion position of the target object;
step 6.3: comparing the size of the target function, and selecting the position point combination of the maximum value of the target function as the selection result of the position points of the two sensors before the array optimization;
step 6.4: and 6.1-6.3 are repeated to finish the selection of the position points of the plurality of sensors.
8. The method for optimizing the position layout of the underwater vehicle side line detection sensors according to claim 6, is characterized in that specifically, by arranging a plurality of sensors on one side of the distribution model, the number of the sensors is gradually increased while the information entropy change is ensured to be maximum, and one sensor distribution position point is selected at a time to determine the optimized overall layout of the sensors.
9. The method for optimizing the position layout of the side line detection sensors of the underwater vehicle as claimed in claim 8, wherein a sensor position point with the largest information entropy reduction is firstly selected as the optimal position S of the first sensor under the condition of having the initial layout of the sensors 1 Passing the optimum position S of the first sensor on the basis of the optimum position of the first sensor 1 The information entropy change obtained by combining the position of the second sensor is calculated, and the sensor combination with the maximum change, namely the maximum objective function, is the first twoOptimum position S of sensor 1 And S 2 Obtaining the optimal distribution combination (S) of the sensors 1 ,S 2 )。
10. An underwater vehicle side line detection sensor position layout optimization system, comprising:
the model establishing module is used for establishing an underwater vehicle physical model and setting a flow field environment of the underwater vehicle physical model;
the theoretical pressure generation module is used for setting initial layout observation points of the sensors according to the physical model of the underwater vehicle and the flow field environment of the underwater vehicle, and performing flow field simulation according to the initial layout observation points to generate a theoretical pressure data set;
the target function optimization module is used for representing a real pressure measurement value according to the obtained theoretical pressure data set and the prediction error, representing a probability density function of the likelihood function, calculating relative entropy of the likelihood function, taking the relative entropy of the prior probability distribution and the posterior probability distribution of the predicted target object motion position and the expected values of all the relative entropies as the target function, converting the unknown quantity in the target function to obtain a converted target function, comparing the size of the converted target function to obtain a first optimization point, sequentially calculating through a sequential heuristic array optimization algorithm to obtain an array layout optimization position coordinate point, optimizing the distribution number of the sensor arrays by calculating the Spanish scale correlation coefficient of each arrangement to determine the optimal number of the sensors, and accordingly completing the optimization of the position layout of the underwater vehicle lateral line detection sensors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110603204.4A CN113361087B (en) | 2021-05-31 | 2021-05-31 | Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110603204.4A CN113361087B (en) | 2021-05-31 | 2021-05-31 | Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113361087A CN113361087A (en) | 2021-09-07 |
| CN113361087B true CN113361087B (en) | 2022-12-09 |
Family
ID=77530668
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110603204.4A Active CN113361087B (en) | 2021-05-31 | 2021-05-31 | Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113361087B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114964078A (en) * | 2022-04-12 | 2022-08-30 | 上海交通大学 | Distribution arrangement method for microwave multi-dimensional deformation and vibration measurement transceivers |
| CN118089747A (en) * | 2024-04-28 | 2024-05-28 | 中国人民解放军军事科学院国防科技创新研究院 | Bionic side line positioning method and system for constant volume target |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902801B (en) * | 2019-01-22 | 2020-11-17 | 华中科技大学 | Flood collective forecasting method based on variational reasoning Bayesian neural network |
| CN111546379B (en) * | 2020-04-26 | 2021-07-13 | 西安交通大学 | Sensor layout method for monitoring operating state of industrial robot |
| CN112163373A (en) * | 2020-09-23 | 2021-01-01 | 中国民航大学 | Radar system performance index dynamic evaluation method based on Bayesian machine learning |
-
2021
- 2021-05-31 CN CN202110603204.4A patent/CN113361087B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113361087A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113361087B (en) | Method and system for optimizing position layout of lateral line detection sensor of underwater vehicle | |
| CN113391357B (en) | Underwater multi-physical-field composite detection system and detection array optimization method | |
| US7342536B2 (en) | System and method for determining the location of emitters having well-behaved timing features | |
| CN106886980B (en) | Point cloud density enhancement method based on three-dimensional laser radar target identification | |
| CN104834994A (en) | Small sample relay protection reliability parameter estimation method based on SVM (Support Vector Machine) | |
| CN105765562B (en) | Method and device for obtaining a data-based function model | |
| CN116609776B (en) | Star convex expansion target tracking method based on artificial potential field method in complex environment | |
| CN115081677B (en) | Wind speed prediction method and device, electronic equipment and readable storage medium | |
| CN114896564A (en) | Transient electromagnetic two-dimensional Bayesian inversion method adopting self-adaptive Thiessen polygon parameterization | |
| EP3779808A1 (en) | Device and method for training a gaussian process regression model | |
| CN117171873B (en) | Vehicle aerodynamic optimization method and device and vehicle | |
| CN104915478A (en) | Product design model equivalent simplifying method based on multi-parameter uncertainty analysis | |
| CN117288365A (en) | Shot peening strengthening residual stress measuring method and device based on neural network | |
| CN114434804B (en) | Bridge-type structure high-precision printing method based on Gaussian process and shape compensation | |
| CN115270301A (en) | Ship type optimization rapid method based on maximization expectation improvement algorithm | |
| Xu et al. | Research on monocular vision distance measurement algorithm based on reference target | |
| EP2442279B1 (en) | Design optimization of physical bodies | |
| CN115327503A (en) | Fixed single-station passive positioning method based on Gaussian particle filtering and related device | |
| JP2006047700A (en) | Three-dimensional bathymetric chart display device and method | |
| CN112388628B (en) | Apparatus and method for training a gaussian process regression model | |
| Liu et al. | Monte Carlo multi-robot localization based on grid cells and characteristic particles | |
| CN111954158A (en) | RSS map-based joint filtering indoor single-target tracking method, device and medium | |
| Wang et al. | A moving grid cell based MCL algorithm for mobile robot localization | |
| CN118067130B (en) | Robot high-precision motion planning method and system based on data fusion | |
| CN117405175B (en) | Intelligent marine environment monitoring system |
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 |