Hit probability prediction method and device, computer equipment and storage medium

Abstract

The invention provides a hit probability prediction method and device, computer equipment and a storage medium, and the method comprises the steps: replacing a random number sequence in a white shark optimization algorithm WSO with a chaotic mapping sequence determined by a chaotic mapping function, and carrying out the population optimization of the white shark optimization algorithm WSO, and obtaining an improved white shark optimization algorithm CWSO; a KAN layer and a fuzzy pooling layer are introduced into a convolutional neural network CNN, and a hit probability prediction model of the KCNN underwater launching device is constructed; obtaining influence factor training data, and performing parameter optimization on the KCNN model through an improved white shark optimization algorithm CWSO by adopting the influence factor training data to obtain a CWSO-KCNN model; and obtaining influence factor target data, and inputting the influence factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device. Therefore, the CWSO is obtained by improving the optimization algorithm, and the CWSO is adopted to perform parameter optimization on the KCNN model to obtain the CWSO-KCNN model, so that the prediction precision of the hit probability is effectively improved.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of artificial intelligence technology, and in particular, to a method, apparatus, computer device, and storage medium for predicting hit probability. Background Art

[0002] In the underwater combat command and control system, accurately predicting the hit probability of different underwater launch device attack methods can provide commanders with an intuitive and rapid basis for combat effectiveness evaluation, helping them optimize the underwater launch device strike plan and thereby enhance the underwater combat capability of surface ships.

[0003] In related technologies, the Monte Carlo method is mainly used to evaluate the hit probability of the supercavitating underwater launch device, and simulation experiments are used to deeply analyze the influence of key parameters such as the initial mine-target distance, the initial target side angle and the target speed on the hit probability.

[0004] However, using the existing method, the prediction accuracy of the hit probability is not high. Summary of the Invention

[0005] The embodiments described herein provide a hit probability prediction method, apparatus, computer device, and storage medium that overcome the aforementioned problems.

[0006] In a first aspect, according to the present disclosure, a hit probability prediction method is provided, comprising:

[0007] The chaotic mapping sequence determined by the chaotic mapping function is used to replace the random number sequence in the White Shark Optimization Algorithm WSO to perform population optimization on the White Shark Optimization Algorithm WSO, thereby obtaining an improved White Shark Optimization Algorithm CWSO;

[0008] The KAN layer and fuzzy pooling layer are introduced into the convolutional neural network (CNN) to build a KCNN underwater launcher hit probability prediction model.

[0009] Obtaining influencing factor training data, and using the influencing factor training data to optimize the parameters of the KCNN model using the improved White Shark Optimization Algorithm (CWSO) to obtain a CWSO-KCNN model, wherein the CWSO-KCNN model is used to predict the hit probability of the underwater launch device;

[0010] Obtain influencing factor target data, and input the influencing factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launch device.

[0011] Optionally, the KCNN underwater launching device hit probability prediction model comprises a CNN layer, a KAN layer and a fuzzy pooling layer, the CNN layer comprises a convolution layer, a pooling layer and a full connection layer, and the fuzzy pooling layer is used for sampling processing of a feature map; the KAN layer is constructed according to the Kolmogorov-Arnold representation theorem and is used for classifying features output by the CNN layer.

[0012] Optionally, the chaotic mapping sequence determined by the chaotic mapping function is used to replace a random number sequence in the white shark optimization algorithm WSO to perform population optimization on the white shark optimization algorithm WSO, so as to obtain an improved white shark optimization algorithm CWSO, which comprises the following steps.

[0013] A corresponding chaotic sequence value is generated by using a chaotic mapping function;

[0014] A white shark position is initialized by using the chaotic sequence value;

[0015] The chaotic sequence value is used to replace a random sequence value to update a moving speed of the white shark towards prey, and the chaotic sequence value is used to replace the random sequence value to update a best moving position;

[0016] The chaotic sequence value is used to replace the random sequence value to determine a white shark group behavior, so as to obtain the improved white shark optimization algorithm CWSO.

[0017] Optionally, the initialization result of the white shark position is represented as:

[0018]

[0019] wherein, represents an initial position of an i th white shark individual in a j th dimension; lb j represents a lower bound of a j th dimension search space; ub j represents an upper bound of the j th dimension search space; SCi represents a chaotic sequence value generated by the i th chaotic mapping;

[0020] The updating speed of the white shark moving towards prey is represented as:

[0021]

[0022] wherein, represents a speed vector of the i th white shark in the k th iteration; μ represents an inertia weight, which is used to control an influence of a historical speed on a current speed; t1 and t2 respectively represent speed adjustment parameters; P gbestk represents a global optimal position in the k th iteration; represents a current position of the i th white shark in the k th iteration; represents a historical optimal position of the i th white shark itself;

[0023] The optimal moving position is expressed as:

[0024]

[0025] in, Represents the distance vector between the global optimal position and the current individual position;

[0026] White shark group behavior is expressed as:

[0027]

[0028] Optionally, the influencing factor training data is used to optimize the parameters of the KCNN model using the improved White Shark Optimization Algorithm CWSO to obtain a CWSO-KCNN model, including:

[0029] When the hyperparameters of the KCNN model are optimized by the improved White Shark optimization algorithm CWSO, the chaotic sequence values ​​are used instead of the random sequence values ​​to update the hyperparameters and parameters of the KCNN model in each iteration, and the hyperparameters and parameters of the KCNN model are adjusted according to the error loss of the hyperparameters and parameters of the updated KCNN model relative to the training data of the influencing factors until the preset iteration requirements are met, thereby obtaining the CWSO-KCNN model.

[0030] Optionally, the fuzzy pooling layer performs a fuzzification operation on the input data to convert it into a fuzzy set, aggregates each fuzzy value in the fuzzy set using a fuzzy logic operation, and converts the aggregated fuzzy data into a crisp output.

[0031] Optionally, the influencing factor training data includes: underwater target motion parameters, guidance system accuracy of the underwater launch device, speed and range of the underwater launch device, underwater environment data and electromagnetic environment; the underwater environment data includes: temperature, salinity and depth.

[0032] In a second aspect, according to the present disclosure, a hit probability prediction device is provided, comprising:

[0033] An optimization module is used to use a chaotic mapping sequence determined by a chaotic mapping function to replace the random number sequence in the White Shark Optimization Algorithm WSO to perform population optimization on the White Shark Optimization Algorithm WSO, thereby obtaining an improved White Shark Optimization Algorithm CWSO;

[0034] A construction module is used to introduce the KAN layer and fuzzy pooling layer into the convolutional neural network (CNN) to build a KCNN underwater launcher hit probability prediction model;

[0035] The processing module is configured to acquire influence factor training data, and use the influence factor training data to perform parameter optimization on the KCNN model by using the improved CWSO to obtain a CWSO-KCNN model, and the CWSO-KCNN model is configured to predict the hit probability of the underwater launching device.

[0036] The determining module is configured to acquire influence factor target data, and input the influence factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device.

[0037] In a third aspect, a computer device is provided, which includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the hit probability prediction method in any one of the above embodiments when executing the computer program.

[0038] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program implements the steps of the hit probability prediction method in any one of the above embodiments when executed by a processor.

[0039] The hit probability prediction method provided by the embodiments of the present application uses the chaotic mapping sequence determined by the chaotic mapping function to replace the random number sequence in the white shark optimization algorithm WSO to perform population optimization on the white shark optimization algorithm WSO, and obtains the improved white shark optimization algorithm CWSO; introduces the KAN layer and the fuzzy pooling layer in the convolutional neural network CNN to construct a KCNN hit probability prediction model of the underwater launching device; acquires influence factor training data, and uses the influence factor training data to perform parameter optimization on the KCNN model by using the improved CWSO to obtain a CWSO-KCNN model, and the CWSO-KCNN model is configured to predict the hit probability of the underwater launching device; acquires influence factor target data, and inputs the influence factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device. In this way, the CWSO is obtained by improving the optimization algorithm, and the CWSO-KCNN model is obtained by using the CWSO to perform parameter optimization on the KCNN model, which effectively improves the prediction accuracy of the hit probability.

[0040] The above description is only a summary of the technical solutions of the embodiments of the present application, in order to more clearly understand the technical means of the embodiments of the present application, the embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, characteristics and advantages of the embodiments of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments will be briefly described in the following. It should be noted that the drawings described below only relate to some of the embodiments of the present disclosure, rather than limiting the present disclosure. Among them:

[0042] Figure 1 is a flow diagram of a hit probability prediction method provided by the present disclosure.

[0043] Figure 2 is a structural diagram of a hit probability prediction device provided by the present disclosure.

[0044] Figure 3 is a structural diagram of a computer device provided by the present disclosure.

[0045] It should be noted that the elements in the drawings are schematic and not drawn to scale. DETAILED DESCRIPTION

[0046] In order to make the objects, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions of the embodiments of the present disclosure will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present disclosure, rather than all the embodiments. Based on the described embodiments of the present disclosure, all other embodiments obtained by a person skilled in the art without any inventive effort are also within the scope of protection of the present disclosure.

[0047] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this present subject matter belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. As used herein, the statement that two or more parts are "connected" or "coupled" together will mean that the parts are joined together either directly or through one or more intermediate parts.

[0048] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase "in an embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily all directed to the same embodiments, alternative or alternative embodiments. It will be explicitly understood by a person of ordinary skill in the art that the embodiments described herein can be combined with other embodiments.

[0049] The term "and / or", used in the present document, only describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A existing, A and B existing, and B existing. In addition, the character " / " in the present document generally represents that the front and rear associated objects are in an "or" relationship. Terms such as "first" and "second" are only used to distinguish one component (or part of a component) from another component (or another part of a component).

[0050] In the description of the present application, unless otherwise specified, "a plurality of" means two or more (including two), and similarly, "a plurality of groups" means two or more groups (including two groups).

[0051] In order for those skilled in the art to better understand the scheme of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings.

[0052] Figure 1 is a flowchart of a hit probability prediction method provided by an embodiment of the present disclosure, as shown in Figure 1 The specific process of the hit probability prediction method includes:

[0053] S110, using the chaotic mapping sequence determined by the chaotic mapping function to replace the random number sequence in the white shark optimization algorithm WSO to perform population optimization on the white shark optimization algorithm WSO, and obtaining an improved white shark optimization algorithm CWSO.

[0054] Among them, by determining the chaotic mapping type (such as ten mapping types such as logical mapping and Chebyshev mapping) and related parameters in the CWSO algorithm, and setting the parameters of the WSO algorithm, including the upper and lower limits of the white shark movement speed, the construction factor, etc., the CWSO parameter initialization setting is realized.

[0055] The present embodiment tends to use chaotic sequence (SCi) rather than simple classic random generator rand() by integrating chaotic mapping, so as to realize more rich and dynamic initial position distribution and optimization exploration stage.

[0056] In some embodiments, the chaotic mapping sequence determined by the chaotic mapping function is used to replace the random number sequence in the white shark optimization algorithm WSO to perform population optimization on the white shark optimization algorithm WSO, and an improved white shark optimization algorithm CWSO is obtained, including:

[0057] The corresponding chaotic sequence value is generated by using the chaotic mapping function.

[0058] Among them, the chaotic sequence value can be represented by the following formula (1).

[0059] SCi = abs(chaos(i)[1, rand, searchAgents - 1]) (1)

[0060] In formula (1), abs represents the absolute value function; chaos(i) represents the ith chaotic mapping function; searchAgents represents the population size, i.e., the number of great white shark individuals.

[0061] The great white shark position is initialized by a chaotic sequence value.

[0062] The initialization result of the great white shark position is represented as formula (2) as follows.

[0063]

[0064] In formula (2), represents the initial position of the ith great white shark individual in the jth dimension; lb j represents the lower bound of the jth dimension search space; ub j represents the upper bound of the jth dimension search space; SCi represents the chaotic sequence value generated by the ith chaotic mapping.

[0065] The chaotic sequence value is used to replace the random sequence value to update the speed of the white shark moving towards the prey, and the chaotic sequence value is used to replace the random sequence value to update the best moving position.

[0066] In which, the displacement of the shark at the (k+1)th iteration is updated according to formula (3) as follows.

[0067]

[0068] The standard WSO algorithm relies on two parameters c1 and c2 randomly selected from the interval [0, 1]. However, the method of the present embodiment uses chaotic sequences to replace these random values to determine the best position in each iteration, and the integration of chaotic sequences enables deeper exploration of the search space and enhances the optimization ability of the algorithm.

[0069] Therefore, the subsequent iteration of the ith shark at iteration k+1, i.e., the update speed of the white shark moving towards the prey, is represented by formula (4) as follows.

[0070]

[0071] In formula (3), represents the velocity vector of the ith great white shark at the kth iteration; μ = (n x SCi) + 1, represents the inertia weight, used to control the influence of the historical speed on the current speed; t1 and t2 represent the speed adjustment parameters, respectively; P gbestk represents the global optimal position at the kth iteration; represents the current position of the i-th great white shark at the k-th iteration; Represents the optimal historical position of the i-th great white shark.

[0072] The CWSO algorithm uses chaotic mapping rather than random values ​​to define parameters in each iteration, which improves the efficiency of the algorithm and produces the best possible global solution. The new optimal position, that is, the best moving position, is expressed as follows:

[0073]

[0074] In formula (5), Represents the distance vector between the global optimal position and the current individual position.

[0075] Because the existing specific formula describing the behavior of great white sharks uses a random function (rand) to determine how each shark on the riverbank adjusts its position based on the two leaders with the optimal solution. However, the rand function has defects, such as long-term predictability, periodicity, and correlation between generated numbers, which can cause the CWSO algorithm to fall into a local optimal solution. Therefore, this embodiment uses a chaotic map to replace the rand function. Chaotic behavior has the following characteristics: Non-periodic, unpredictable short-term behavior: Chaotic systems (such as chaotic maps) do not follow fixed short-term cycles, although patterns or structures may appear over a long period of time. Long-term repeatability: When observed over a sufficiently long time scale, some chaotic systems may exhibit repetitive behavior or potential patterns. However, these patterns are not necessarily regular periods like periodic systems. Sensitivity to initial conditions: A fundamental characteristic of chaotic systems is sensitivity to initial conditions, which means that small changes in initial conditions can lead to completely different behaviors, making short-term trajectories unpredictable.

[0076] Chaotic sequence values ​​are used instead of random sequence values ​​to determine the behavior of white shark groups, so as to obtain the improved white shark optimization algorithm CWSO.

[0077] In particular, this embodiment improves the accuracy and precision of modeling the social behavior of great white sharks by incorporating chaos mapping into the model, and optimizes the solution search process in a given space.

[0078] The behavior of white shark groups is expressed as the following formula (6).

[0079]

[0080] S120. Introduce the KAN layer and fuzzy pooling layer into the convolutional neural network (CNN) to construct a KCNN underwater launcher hit probability prediction model.

[0081] The KCNN model is constructed by combining the KAN and the CNN. The KAN uses a learnable activation function represented by a spline based on the Kolmogorov-Arnold representation theorem, and can optimize the learned features. The complex relationship between the features can be further mined based on the local features extracted by the CNN, the expression ability of the model to the features is enhanced, and the generalization performance of the model is improved. In the hit probability prediction of the underwater launching device, the KAN can perform deep analysis on the target and underwater launching device related features extracted by the CNN, and improve the adaptability of the model to different combat scenarios.

[0082] In some embodiments, the hit probability prediction model of the KCNN underwater launching device includes a CNN layer, a KAN layer and a fuzzy pooling layer. The CNN layer includes a convolution layer, a pooling layer and a full connection layer. The number of layers of the full connection layer, the size of the convolution kernel and the step length and the like can be set. The fuzzy pooling layer is used for sampling processing of the feature map. The KAN layer is constructed according to the Kolmogorov-Arnold representation theorem, and is used for classifying the features output by the CNN layer.

[0083] The Kolmogorov-Arnold network can be represented by the following formula (7).

[0084]

[0085] In formula (7), x represents an input vector, including n-dimensional features such as target motion parameters, guidance system accuracy of the underwater launching device and the like; Φ q represents the qth activation function; φ q,p (x p ) represents a unary transformation function of the pth input feature in the qth basis function, which is the sum of the basis function b(x) and the spline function, and can be represented by the following formula (8).

[0086] φ(x)=w b b(x)+w s spline(x) (8)

[0087] In formula (8), spline(x) represents a spline function; w b and w s respectively represent trainable factors, which can better control the overall amplitude of the activation function.

[0088] The basis function b(x) can be represented by the following formula (9).

[0089]

[0090] In equation (9), silu is a Sigmoid linear unit activation function; the spline function spline(x) is parameterized as a linear combination of B-splines, as expressed in equation (10) below.

[0091] spline(x) =∑ i c i B i (x) (10)

[0092] In equation (10), c i represents a trainable factor.

[0093] The depth KAN is expressed in equation (11) below.

[0094]

[0095] In some embodiments, the fuzzy pooling layer converts the input data into fuzzy sets by performing a fuzzification operation, aggregates each fuzzified value in the fuzzy sets using fuzzy logic operations, and converts the aggregated fuzzy data into crisp output.

[0096] The key operations of fuzzy pooling include fuzzification, aggregation, and defuzzification. The fuzzification operation converts crisp input into fuzzy sets, enabling the pooling layer to handle uncertainty in the data. After fuzzification, the fuzzy pooling layer aggregates the fuzzified values within each pooling window using fuzzy logic operations, which involves fuzzy algebra and methods such as max-of-mins, considering the contribution of all elements within the window rather than selecting a single maximum or taking an average. The defuzzification operation converts the aggregated fuzzy data back into crisp output.

[0097] For example, consider a set of three fuzzy sets defined by equation (12) below.

[0098]

[0099] For V = 3, let be the element of volume patch p at position (i,j) for depth n, where i = 1,...,k, k is the size of the pooling window, and the triangular membership function h1 used for patch fuzzification is defined as expressed in equation (13) below.

[0100]

[0101] In equation (13), d = r max / 2; o = d / 3.

[0102] The membership function h2 is defined as expressed in equation (14) below.

[0103]

[0104] In equation (14), a = r max / 4; m = r max / 2; b = m + a.

[0105] The membership function h3 is defined as shown in equation (15).

[0106]

[0107] In equation (15), r = r max / 2; q = r + r max / 4; r max = 6.

[0108] For each patch p n , n = 1,..., z, the blurred patch p is computed as shown in equation (16).

[0109]

[0110] The values within each pooling window are aggregated using fuzzy algebra and the fuzzy values are aggregated as shown in equation (17).

[0111]

[0112] In equation (17), p is the fraction that quantifies the overall membership of the pair. Based on these fractions for each patch p , a new fuzzy volume patch p is created by selecting the spatial fuzzy patch p with the largest fraction v = 1,..., V as shown in equation (18).

[0113]

[0114] After selecting the patch with the highest certainty, the dimension of each patch is reduced by deblurring using the center of gravity method (COG) as shown in equation (19).

[0115]

[0116] In equation (19), p' = {p'n | n = 1, 2,..., z}.

[0117] In this embodiment, the CNNs combine a fuzzy pooling layer to sample the feature maps, and use KAN as the classification head in the last layer of the network, instead of the traditional MLP (Multilayer Perceptron). The ability of fuzzy pooling to extract valuable features from uncertain and noisy feature maps, combined with the adaptability of KAN, makes the model more efficient in terms of memory usage, as KAN can achieve comparable performance to MLP with fewer parameters.

[0118] In S130, the influence factor training data is obtained, and the CWSO-KCNN model is obtained by optimizing the parameters of the KCNN model through the improved CWSO algorithm based on the influence factor training data, and the CWSO-KCNN model is used to predict the hit probability of the underwater launching device.

[0119] In some embodiments, the influence factor training data includes underwater target motion parameters, guidance system accuracy of the underwater launching device, speed and range of the underwater launching device, underwater environment data, and electromagnetic environment; the underwater environment data includes temperature, salinity, and depth. Specifically, the influence factor training data includes training data of underwater target motion parameters such as azimuth, speed, and heading; training data of guidance system accuracy of the underwater launching device, training data of speed and range of the underwater launching device, training data of underwater environment data, and training data of electromagnetic environment; the underwater environment data includes sea temperature, salinity, and depth.

[0120] In some embodiments, the CWSO-KCNN model is obtained by optimizing the parameters of the KCNN model through the improved CWSO algorithm based on the influence factor training data, and the CWSO-KCNN model is obtained, including:

[0121] When the hyperparameters of the KCNN model are optimized through the improved CWSO algorithm, in each iteration, the hyperparameters and parameters of the KCNN model are updated by using chaotic sequence values instead of random sequence values, and the hyperparameters and parameters of the KCNN model are adjusted according to the error loss of the updated KCNN model relative to the influence factor training data, until the preset iteration requirement is met, and the CWSO-KCNN model is obtained.

[0122] Wherein, the parameters (such as KAN layer spline function coefficients, CNN layer convolution kernel weights) and hyperparameters (learning rate, training rounds) of the KCNN model are optimized through the CWSO algorithm using the training set data. In each iteration, according to the rules of the CWSO algorithm, the model parameters are updated by using chaotic sequence instead of random numbers, the error loss (such as cross-entropy loss function) of the model on the training set is calculated, and the parameters are adjusted according to the error, until the preset training rounds or error convergence condition is reached.

[0123] In addition, the model in the training process can be evaluated by the verification set to monitor the performance indicators (such as accuracy, recall rate, and F1 value) of the model. If the performance of the model on the verification set no longer improves or shows signs of overfitting (such as the verification set error starts to rise), the training strategy is adjusted (such as stopping training early or adjusting the learning rate). After completing the training, the model is comprehensively evaluated using the test set, and the performance of the KCNN model enhanced by CWSO is compared with that of other comparative models. The evaluation indicators include accuracy, convergence speed (observing the change of error loss function with the number of iterations), and stability (calculating the standard deviation of multiple experimental results).

[0124] S140, obtaining influence factor target data, and inputting the influence factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device.

[0125] The influence factor target data includes test data of underwater target motion parameters such as azimuth, speed, and heading; test data of the guidance system accuracy of the underwater launching device, test data of the speed and range of the underwater launching device, test data of underwater environmental data, and test data of electromagnetic environment; and underwater environmental data including sea temperature, salinity, and depth.

[0126] The hit probability prediction value of the underwater launching device can be represented by a specific probability value, such as 90%, to visually demonstrate the effectiveness of the CWSO-KCNN model.

[0127] In this embodiment, the chaotic mapping sequence determined by the chaotic mapping function is used to replace the random number sequence in the white shark optimization algorithm WSO to optimize the population of the white shark optimization algorithm WSO, obtaining an improved white shark optimization algorithm CWSO; the KAN layer and the fuzzy pooling layer are introduced into the convolutional neural network CNN to construct a KCNN hit probability prediction model for the underwater launching device; influence factor training data is obtained, and the influence factor training data is used to optimize the parameters of the KCNN model by the improved white shark optimization algorithm CWSO to obtain a CWSO-KCNN model, which is used to predict the hit probability of the underwater launching device; influence factor target data is obtained, and the influence factor target data is input into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device. In this way, the CWSO is obtained by improving the optimization algorithm, and the CWSO-KCNN model is obtained by optimizing the parameters of the KCNN model using the CWSO, effectively improving the prediction accuracy of the hit probability of the underwater launching device.

[0128] Figure 2 A structure schematic diagram of a hit probability prediction device provided in this embodiment can include an optimization module 210, a construction module 220, a processing module 230, and a determination module 240.

[0129] wherein:

[0130] The optimization module 210 is configured to replace the random number sequence in the white shark optimization algorithm WSO with the chaotic mapping sequence determined by the chaotic mapping function, to perform population optimization on the white shark optimization algorithm WSO, and to obtain an improved white shark optimization algorithm CWSO.

[0131] The construction module 220 is configured to introduce a KAN layer and a fuzzy pooling layer into a convolutional neural network CNN, and to construct a KCNN underwater launching device hit probability prediction model.

[0132] The processing module 230 is configured to obtain influence factor training data, and to perform parameter optimization on the KCNN model by using the influence factor training data and the improved white shark optimization algorithm CWSO to obtain a CWSO-KCNN model, which is configured to predict the hit probability of the underwater launching device.

[0133] The determination module 240 is configured to obtain influence factor target data, and to input the influence factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launching device.

[0134] In this embodiment, optionally, the KCNN underwater launching device hit probability prediction model comprises a CNN layer, a KAN layer and a fuzzy pooling layer, the CNN layer comprises a convolutional layer, a pooling layer and a full connection layer, and the fuzzy pooling layer is configured to sample and process feature maps; the KAN layer is constructed according to the Kolmogorov-Arnold representation theorem, and is configured to classify features output by the CNN layer.

[0135] In this embodiment, optionally, the optimization module 210 is specifically configured to:

[0136] generate a corresponding chaotic sequence value by using the chaotic mapping function; initialize the position of the great white shark by using the chaotic sequence value; update the speed of the white shark moving towards the prey by using the chaotic sequence value instead of the random sequence value, and update the best moving position by using the chaotic sequence value instead of the random sequence value; determine the group behavior of the white shark by using the chaotic sequence value instead of the random sequence value, to obtain the improved white shark optimization algorithm CWSO.

[0137] In this embodiment, optionally, the initialization result of the position of the great white shark is represented as:

[0138]

[0139] wherein, represents the initial position of the i th great white shark individual in the j th dimension; lb j represents the lower bound of the j th dimension search space; ub jUpper bound of the jth dimension search space; SCi represents the chaotic sequence value generated by the ith chaotic mapping.

[0140] The update speed of the white shark moving towards the prey is represented as:

[0141]

[0142] wherein, vi(k) represents the velocity vector of the ith white shark at the kth iteration; μ represents the inertia weight for controlling the influence of the historical velocity on the current velocity; t1 and t2 represent the velocity adjustment parameters, respectively; P gbestk xk represents the global optimal position at the kth iteration; xi(k) represents the current position of the ith white shark at the kth iteration; xi(k) represents the current position of the ith white shark at the kth iteration;

[0143] The optimal moving position is represented as:

[0144]

[0145] wherein, dxi(k) represents the distance vector between the global optimal position and the current individual position.

[0146] The white shark swarm behavior is represented as:

[0147]

[0148] In this embodiment, the processing module 230 is specifically configured to:

[0149] When the improved white shark optimization algorithm CWSO is used to optimize the hyperparameters of the KCNN model, the hyperparameters and parameters of the KCNN model are updated by using chaotic sequence values instead of random sequence values in each iteration, and the hyperparameters and parameters of the KCNN model are adjusted according to the error loss of the updated hyperparameters and parameters of the KCNN model relative to the influence factor training data until the preset iteration requirement is met, and the CWSO-KCNN model is obtained.

[0150] In this embodiment, the fuzzy pooling layer converts the input data into a fuzzy set by performing a fuzzy operation, aggregates each fuzzy value in the fuzzy set using fuzzy logic operations, and converts the aggregated fuzzy data into clear output.

[0151] In this embodiment, the influence factor training data includes underwater target motion parameters, underwater launch device guidance system accuracy, underwater launch device speed and range, underwater environmental data, and electromagnetic environment; the underwater environmental data includes temperature, salinity, and depth.

[0152] The hit probability prediction apparatus provided by the present disclosure can execute the method embodiments, and the specific implementation principles and technical effects can be referred to the method embodiments, which will not be described here again.

[0153] The present application also provides a computer device. For details, please refer to Figure 3 , Figure 3 The basic structure block diagram of the computer device is shown in the figure.

[0154] The computer device includes a memory 310 and a processor 320 which are connected to each other through a system bus. It should be noted that only the computer device with the memory 310 and the processor 320 is shown in the figure, but it should be understood that it is not required to implement all the shown components, and more or less components can be alternatively implemented. Among them, the computer device herein is a device capable of automatically performing numerical calculation and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0155] The computer device can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing devices. The computer device can interact with the user through a keyboard, a mouse, a remote controller, a touchpad or a voice control device.

[0156] The memory 310 includes at least one type of readable storage medium, including non-volatile memory or volatile memory, for example, flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. The RAM can include static RAM or dynamic RAM. In some embodiments, the memory 310 can be an internal storage unit of the computer device, for example, a hard disk or a memory of the computer device. In other embodiments, the memory 310 can also be an external storage device of the computer device, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, or a flash card, etc. equipped on the computer device. Of course, the memory 310 can include both an internal storage unit and an external storage device of the computer device. In the present embodiment, the memory 310 is generally used to store an operating system and various application software installed on the computer device, for example, program codes of the above-described method, etc. In addition, the memory 310 can also be used to temporarily store various data that has been output or will be output.

[0157] The processor 320 is generally used to perform the overall operation of the computer device. In the present embodiment, the memory 310 is used to store program codes or instructions, which include computer operation instructions, and the processor 320 is used to execute the program codes or instructions stored in the memory 310 or process data, for example, run the program codes of the above-described method.

[0158] In this article, the bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus system can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is shown in the figure, but it does not mean that there is only one bus or only one type of bus.

[0159] Another embodiment of the present application also provides a computer readable medium, which can be a computer readable signal medium or a computer readable medium. The processor in the computer reads the computer readable program code stored in the computer readable medium, so that the processor can perform the function actions specified in each step or combination of steps in the above method; generate the device implementing the function actions specified in each block or combination of blocks in the block diagram.

[0160] The computer readable medium includes but is not limited to electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any appropriate combination of the foregoing, for storing program code or instructions, which include computer operation instructions, and the processor for executing the program code or instructions of the above method stored in the memory.

[0161] The definition of the memory and the processor can refer to the description of the foregoing computer device embodiment, which will not be repeated here.

[0162] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiment described above is only schematic, for example, the division of modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be through some interface, indirect coupling or communication connection between the units or devices, which can be electrical, mechanical or other forms.

[0163] The function units or modules in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software function unit.

[0164] If the integrated unit is implemented in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or say the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0165] In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The application can be implemented by means of both hardware and software, and any combination thereof. In the device claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage. The use of relative terms such as "first", "second" and "third", etc. does not connote any prioritization, but such terms are used to distinguish a certain feature from another feature with the same name. The steps of the methods described in the above embodiments should not be understood as necessarily limited in their sequence, except when this is explicitly specified.

[0166] The above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; even though the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for predicting hit probability, characterized in that: include: The chaotic mapping sequence determined by the chaotic mapping function is used to replace the random number sequence in the White Shark Optimization Algorithm WSO to perform population optimization on the White Shark Optimization Algorithm WSO, thereby obtaining an improved White Shark Optimization Algorithm CWSO; The KAN layer and fuzzy pooling layer are introduced into the convolutional neural network (CNN) to build a KCNN underwater launcher hit probability prediction model. Obtaining influencing factor training data, and using the influencing factor training data to optimize the parameters of the KCNN model using the improved White Shark Optimization Algorithm (CWSO) to obtain a CWSO-KCNN model, wherein the CWSO-KCNN model is used to predict the hit probability of the underwater launch device; Obtain influencing factor target data, and input the influencing factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launch device.

2. The method according to claim 1, characterized in that The KCNN underwater launch device hit probability prediction model includes: a CNN layer, a KAN layer and a fuzzy pooling layer. The CNN layer includes: a convolution layer, a pooling layer and a fully connected layer. The fuzzy pooling layer is used to sample the feature map; the KAN layer is constructed based on the Kolmogorov-Arnold representation theorem and is used to classify the features output by the CNN layer.

3. The method according to claim 1, characterized in that The chaotic mapping sequence determined by the chaotic mapping function is used to replace the random number sequence in the White Shark Optimization Algorithm WSO to perform population optimization on the White Shark Optimization Algorithm WSO, thereby obtaining an improved White Shark Optimization Algorithm CWSO, including: Adopting chaotic mapping function to generate corresponding chaotic sequence value; Initializing the great white shark position through the chaotic sequence value; Using the chaotic sequence value instead of the random sequence value to update the speed at which the white shark moves toward the prey, and using the chaotic sequence value instead of the random sequence value to update the optimal moving position; The chaotic sequence value is used instead of the random sequence value to determine the white shark group behavior, so as to obtain the improved white shark optimization algorithm CWSO.

4. The method according to claim 3, characterized in that The initialization result of the great white shark position is expressed as: Among them, P j i represents the initial position of the i-th great white shark individual in the j-th dimension; lb j represents the lower bound of the j-th dimension search space; ub j represents the upper bound of the j-th dimension search space; SCi represents the chaotic sequence value generated by the i-th chaotic map; The updated speed of the white shark moving towards its prey is expressed as: Among them, v k i represents the velocity vector of the i-th great white shark at the k-th iteration; μ represents the inertia weight, which is used to control the influence of historical speed on current speed; t1 and t2 represent speed adjustment parameters respectively; P gbestk represents the global optimal position at the kth iteration; P k i represents the current position of the i-th great white shark at the k-th iteration; represents the optimal historical position of the i-th great white shark; The optimal moving position is expressed as: in, Represents the distance vector between the global optimal position and the current individual position; White shark group behavior is expressed as:

5. The method according to claim 4, characterized in that The influencing factor training data is used to optimize the parameters of the KCNN model using the improved White Shark Optimization Algorithm CWSO to obtain the CWSO-KCNN model, including: When the hyperparameters of the KCNN model are optimized by the improved White Shark optimization algorithm CWSO, the chaotic sequence values ​​are used instead of the random sequence values ​​to update the hyperparameters and parameters of the KCNN model in each iteration, and the hyperparameters and parameters of the KCNN model are adjusted according to the error loss of the hyperparameters and parameters of the updated KCNN model relative to the training data of the influencing factors until the preset iteration requirements are met, thereby obtaining the CWSO-KCNN model.

6. The method according to claim 2, characterized in that The fuzzy pooling layer performs fuzzification operations on the input data, converts it into a fuzzy set, aggregates each fuzzy value in the fuzzy set using fuzzy logic operations, and converts the aggregated fuzzy data into a crisp output.

7. The method according to claim 1, characterized in that The influencing factor training data includes: underwater target motion parameters, the guidance system accuracy of the underwater launch device, the speed and range of the underwater launch device, underwater environment data and electromagnetic environment; the underwater environment data includes: temperature, salinity and depth.

8. A hit probability prediction device, characterized in that: include: An optimization module is used to use a chaotic mapping sequence determined by a chaotic mapping function to replace the random number sequence in the White Shark Optimization Algorithm WSO to perform population optimization on the White Shark Optimization Algorithm WSO, thereby obtaining an improved White Shark Optimization Algorithm CWSO; A construction module is used to introduce the KAN layer and fuzzy pooling layer into the convolutional neural network (CNN) to build a KCNN underwater launcher hit probability prediction model; a processing module for obtaining influencing factor training data, and using the influencing factor training data to optimize the parameters of the KCNN model using the improved White Shark Optimization Algorithm (CWSO) to obtain a CWSO-KCNN model, wherein the CWSO-KCNN model is used to predict the hit probability of the underwater launch device; The determination module is used to obtain influencing factor target data and input the influencing factor target data into the CWSO-KCNN model to obtain a hit probability prediction value of the underwater launch device.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the method for predicting the hit probability as claimed in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the hit probability prediction method according to any one of claims 1 to 7 is implemented.

Images