Unmanned ship navigation control method and device based on emotion factors, equipment and medium

Abstract

The invention discloses an unmanned ship navigation control method and device based on an emotion factor, equipment and a medium, and relates to the technical field of unmanned ship control, and the method comprises the steps: obtaining the state information of an unmanned ship and the environment information of the environment where the unmanned ship is located, and constructing a target vector based on the state information and the environment information of the unmanned ship; inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model; inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain a navigation instruction output by the navigation decision model; and controlling the unmanned ship to sail based on the sailing instruction. The method and the device are used for solving the problems of poor flexibility and poor adaptability when the unmanned ship is controlled to operate autonomously by using a manually set rule in the prior art, the navigation decision matched with the state information and the environment information of the unmanned ship is automatically generated, and the flexibility, the adaptability and the accuracy of autonomous operation of the unmanned ship are improved.

Description

Technical Field

[0001] The present application relates to the technical field of unmanned ship control, and in particular to an unmanned ship navigation control method, device, equipment and medium based on emotion factors. Background Art

[0002] Unmanned ships are highly information-based ships that can sail automatically by carrying various sensors and autonomous control systems, effectively replacing the traditional manual operation method. The application of unmanned ships has experienced a development from simple scenarios to complex scenarios. Initially, unmanned ships were used in military exercises in the form of remote-controlled target ships. With the improvement of the degree of autonomy, unmanned ships have been widely used in various operation scenarios, such as cruise surveillance, target tracking and hydrological monitoring.

[0003] Accurate navigation decision-making is the key to autonomous navigation of unmanned ships. In order to achieve autonomous navigation of unmanned ships, existing technologies use experience-based rules to generate navigation decisions, and based on these rules, navigation decisions that control the operation of unmanned ships are obtained. However, this method relies on artificially set rules, has poor flexibility, and is difficult to adapt to navigation processes in a wide range of scenarios. Summary of the invention

[0004] In response to the above-mentioned problems and technical needs, the applicant has proposed an unmanned ship navigation control method, device, equipment and medium based on emotional factors to solve the problems of poor flexibility and adaptability in the prior art when using artificially set rules to control the autonomous operation of unmanned ships, and to automatically generate navigation decisions that match the unmanned ship's state information and environmental information, thereby improving the flexibility, adaptability and accuracy of the unmanned ship's autonomous operation.

[0005] The embodiment of the present application provides a method for controlling the navigation of an unmanned ship based on emotional factors, the method comprising:

[0006] Acquire the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and construct a target vector based on the state information of the unmanned ship and the environmental information;

[0007] Inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on target vector samples and emotion factor samples;

[0008] Inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain a navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample;

[0009] Based on the navigation instruction, the unmanned ship is controlled to navigate.

[0010] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, before acquiring the unmanned ship state information and the environmental information of the environment where the unmanned ship is located, the method further includes:

[0011] Acquire a route to be navigated corresponding to the unmanned ship, wherein the route to be navigated includes: at least two navigation nodes and route indication information between any two navigation nodes;

[0012] After obtaining the environmental information of the unmanned ship's environment, it also includes:

[0013] Based on the unmanned ship state information, determining a current navigation node corresponding to the unmanned ship;

[0014] Based on the to-be-navigated route and the current navigation node, determining the next navigation node corresponding to the unmanned ship, and target route indication information between the current navigation node and the next navigation node;

[0015] The step of controlling the unmanned ship to sail based on the navigation instruction includes:

[0016] Based on the navigation instruction and the target route indication information, the unmanned ship is controlled to navigate.

[0017] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, the navigation instruction includes: the navigation speed of the unmanned ship and the navigation direction of the unmanned ship;

[0018] The target route indication information includes: the target route indication direction;

[0019] The step of controlling the navigation of the unmanned ship based on the navigation instruction and the target route indication information includes:

[0020] Based on the navigation direction and the target route indication direction, determining whether the unmanned ship deviates;

[0021] When it is determined that the unmanned ship does not deviate, controlling the navigation of the unmanned ship based on the navigation speed and the navigation direction;

[0022] When it is determined that the unmanned ship deviates, the navigation of the unmanned ship is controlled based on the navigation speed and the target route indication direction.

[0023] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, the neural network model includes: a self-organizing map model;

[0024] The step of inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model includes:

[0025] The target vector is input into the self-organizing map model, the target vector is mapped into a multidimensional grid structure through the self-organizing map model, a winning neuron is determined based on a preset expected target and a preset evaluation function, and the emotion factor is output based on the winning neuron.

[0026] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, before inputting the target vector and the emotional factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, the method further includes:

[0027] Each parameter in the emotion factor is normalized to obtain a normalized emotion factor.

[0028] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, the training process of the navigation decision model includes:

[0029] Acquire the target vector sample, the emotion factor sample and the navigation instruction sample;

[0030] Inputting the target vector sample and the emotion factor sample into the navigation decision model to obtain a predicted navigation instruction output by the navigation decision model;

[0031] Obtaining a predicted reward based on the predicted navigation instruction, the navigation instruction sample and a preset reward function;

[0032] Based on the predicted reward, the model parameters of the navigation decision model are adjusted until the predicted reward is continuously maintained within a first preset range, and it is determined that the training of the navigation decision model is completed.

[0033] According to an unmanned ship navigation control method based on emotional factors in one embodiment of the present application, the training process of the self-organizing map model includes:

[0034] Acquire the target vector sample and the emotion factor sample;

[0035] Input the target vector sample into the self-organizing map model, map the target vector sample into a multidimensional grid structure through the self-organizing map model, determine the winning neuron based on a preset expected target and a preset evaluation function, and output a predicted emotion factor based on the winning neuron;

[0036] Comparing the consistency of the predicted emotion factor and the emotion factor sample to obtain a comparison result;

[0037] Based on the comparison result, the model parameters of the self-organizing map model are adjusted until the comparison result is continuously maintained within a second preset range, and it is determined that the training of the self-organizing map model is completed.

[0038] The embodiment of the present application also provides an unmanned ship navigation control device based on emotion factors, including:

[0039] An acquisition module, used to acquire the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and to construct a target vector based on the state information of the unmanned ship and the environmental information;

[0040] A processing module, used for inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotion factor sample;

[0041] A decision module, used for inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain a navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample;

[0042] A control module is used to control the navigation of the unmanned ship based on the navigation instruction.

[0043] An embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the unmanned ship navigation control method based on emotional factors as described in any one of the above items are implemented.

[0044] An embodiment of the present application also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the unmanned ship navigation control method based on emotional factors as described in any of the above items are implemented.

[0045] The unmanned ship navigation control method, device, equipment and medium based on emotional factors provided in the embodiments of the present application obtain the state information of the unmanned ship and the environmental information of the environment in which the unmanned ship is located, and obtain the corresponding target vector; the target vector is input into a pre-trained neural network model to obtain the emotional factor output by the neural network model, and the emotional factor is used to characterize the difficulty of the current navigation of the unmanned ship. It can be seen that the present application determines the current actual driving condition and the current driving difficulty, and provides an effective data basis for determining subsequent navigation decisions; then, the target vector and the emotional factor are input into the navigation decision model to obtain the navigation instruction output by the navigation decision model; based on the navigation instruction, the navigation of the unmanned ship is controlled. Based on the actual driving condition and the difficulty of driving, the present application uses the navigation decision model to quickly and accurately obtain a matching navigation strategy (corresponding to the navigation instruction) to perform autonomous control of the unmanned ship.

[0046] The present invention is not limited to the state of the unmanned ship and the environment in which the unmanned ship is located. It generates a navigation strategy that matches the actual driving condition and driving difficulty of the unmanned ship. It has high flexibility and adaptability, and effectively solves the problems of poor flexibility and adaptability that occur when the prior art uses artificially set rules to control the autonomous operation of unmanned ships. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 This is one of the flow charts of the unmanned ship navigation control method based on emotion factors provided in the embodiment of the present application;

[0049] Figure 2 It is a structural diagram of a self-organizing map model provided in an embodiment of the present application;

[0050] Figure 3 This is the second flow chart of the unmanned ship navigation control method based on emotion factors provided in the embodiment of the present application;

[0051] Figure 4 is a structural schematic diagram of an unmanned ship navigation control device based on emotion factors provided in an embodiment of the present application;

[0052] Figure 5 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] In order to clearly explain this application, the navigation control of the unmanned ship in the ocean scene is further explained:

[0055] In the prior art, there is a method of using reinforcement learning to generate navigation decisions, specifically by obtaining a large number of data samples and learning directly from the data samples to improve the adaptability of navigation decision generation. However, the main disadvantage of reinforcement learning is that it is difficult to balance the "exploration-utilization" relationship, especially for the navigation decision-making task of unmanned ships. Due to the large number of input state quantities and constraints that need to be considered, the model training efficiency is often low due to insufficient utilization of data samples, and the navigation decision accuracy of the final model output is poor.

[0056] This application is aimed at navigation scenarios with multiple scenarios and multiple constraints. Conventional reinforcement learning algorithms have poor utilization of sample data and are difficult to converge effectively.

[0057] The specific implementation of the present application is further described below in conjunction with the accompanying drawings.

[0058] The embodiment of the present application provides a method for controlling the navigation of an unmanned ship based on emotional factors. The method can be applied to a smart terminal, a server, or a controller of an unmanned ship. Some other descriptions in the embodiment of the present application are for illustrative purposes only and are not intended to limit the scope of protection of the present application. They will not be described one by one. The specific implementation of the method is as follows: Figure 1 As shown:

[0059] Step 101, obtain the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and construct a target vector based on the state information of the unmanned ship and the environmental information.

[0060] Among them, the unmanned ship status information includes: the current position, current speed, current heading and other information of the unmanned ship; the environmental information includes: the position of the current navigation node, the position of the next navigation node, sea obstacles within a preset range, and related information of sea obstacles and other information.

[0061] Among them, the relevant information of marine obstacles includes the type, size, shape, whether the obstacle is moving, the speed of movement, the direction of movement, etc.

[0062] Specifically, the status information of the unmanned ship is obtained through the ship-borne inertial navigation and satellite navigation system (Global Positioning System, GPS for short); the environmental information is obtained through electronic nautical charts, ship automatic identification system (AIS) and ship-borne real-time detection sensors.

[0063] Step 102, input the target vector into the pre-trained neural network model to obtain the emotion factor output by the neural network model.

[0064] Among them, the emotional factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector samples and the emotional factor samples.

[0065] The emotion factors are displayed in the form of an array, for example, S = [α, β, γ, δ, ε].

[0066] Among them, S represents the emotional factor, α represents fear, β represents anger, γ represents surprise, δ represents happiness, and ε represents sadness.

[0067] Among them, the neural network model includes: a self-organizing map model.

[0068] Step 103, input the target vector and the emotion factor into the pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model.

[0069] Among them, the navigation decision model is trained based on target vector samples, emotion factor samples and navigation instruction samples.

[0070] Among them, the navigation instructions include the navigation speed and navigation direction of the unmanned ship.

[0071] Step 104, based on the navigation instruction, control the unmanned ship to navigate.

[0072] Specifically, the unmanned ship is controlled to sail according to the sailing speed and sailing direction in the sailing instructions.

[0073] Specifically, this application needs to obtain the status information of the unmanned ship and the environmental information of the environment in which the unmanned ship is located in real time, so as to generate navigation instructions based on this, so as to control the navigation of the unmanned ship in real time according to actual conditions and ensure the safety of the unmanned ship's navigation.

[0074] The unmanned ship navigation control method based on emotion factors provided in the embodiment of the present application obtains the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and obtains the corresponding target vector; the target vector is input into a pre-trained neural network model to obtain the emotion factor output by the neural network model, and the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship. It can be seen that the present application determines the current actual driving condition and the current driving difficulty, and provides an effective data basis for determining subsequent navigation decisions; then, the target vector and the emotion factor are input into the navigation decision model to obtain the navigation instruction output by the navigation decision model; based on the navigation instruction, the navigation of the unmanned ship is controlled. Based on the actual driving condition and the difficulty of driving, the present application uses the navigation decision model to quickly and accurately obtain a matching navigation strategy (corresponding to the navigation instruction) to perform autonomous control of the unmanned ship.

[0075] The present invention is not limited to the state of the unmanned ship and the environment in which the unmanned ship is located. It generates a navigation strategy that matches the actual driving condition and driving difficulty of the unmanned ship. It has high flexibility and adaptability, and effectively solves the problems of poor flexibility and adaptability that occur when the prior art uses artificially set rules to control the autonomous operation of unmanned ships.

[0076] In a specific embodiment, before obtaining the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, the route to be navigated corresponding to the unmanned ship is obtained. After obtaining the environmental information of the environment where the unmanned ship is located, the current navigation node corresponding to the unmanned ship is determined based on the state information of the unmanned ship; based on the route to be navigated and the current navigation node, the next navigation node corresponding to the unmanned ship and the target route indication information between the current navigation node and the next navigation node are determined. Then, based on the navigation instruction and the target route indication information, the navigation of the unmanned ship is controlled.

[0077] The route to be navigated includes: at least two navigation nodes and route indication information between any two navigation nodes. The route to be navigated also includes the node position of each navigation node.

[0078] The target route indication information includes: the target route indication direction.

[0079] Specifically, before sailing, the unmanned ship obtains the route to be sailed in advance, and departs from the starting point to the end point based on the navigation nodes in the route to be sailed.

[0080] Specifically, the current navigation node corresponding to the unmanned ship is determined according to the current position of the unmanned ship and the node position of the navigation node in the unmanned ship state information.

[0081] In a specific embodiment, based on the navigation instruction and the target route indication information, the specific implementation of controlling the navigation of the unmanned ship includes:

[0082] Based on the navigation direction and the target route indication direction, determine whether the unmanned ship deviates; if it is determined that the unmanned ship does not deviate, control the navigation of the unmanned ship based on the navigation speed and navigation direction; if it is determined that the unmanned ship deviates, control the navigation of the unmanned ship based on the navigation speed and the target route indication direction.

[0083] Specifically, a matching range between the navigation direction and the target route indication direction is preset, that is, if the navigation direction belongs to the corresponding range with the target route execution direction, it is determined that the unmanned ship is not yawed, otherwise, it is determined that the unmanned ship is yawed.

[0084] In a specific embodiment, the specific implementation of processing the state information and environment information of the unmanned ship based on the preset data processing strategy to obtain the target vector and the emotion factor includes:

[0085] The state information and environmental information of the unmanned ship are vectorized to obtain a target vector; the target vector is input into a pre-trained self-organizing map model to obtain the emotion factor output by the self-organizing map model.

[0086] Among them, the self-organizing map network is trained based on target vector samples and emotion factor samples.

[0087] Specifically, the unmanned ship state information and environment information can be vectorized using existing methods, such as the One Hot Model, Bag of Words Model, N-Gram, Word2vec, and Doc2vec. The unmanned ship state information and environment information are vectorized based on any one or more of the above models to obtain a target vector.

[0088] In a specific embodiment, the specific implementation of inputting the target vector into a pre-trained neural network model to obtain the emotion factor output by the neural network model includes:

[0089] The target vector is input into the self-organizing map model, and the target vector is mapped to the multidimensional grid structure through the self-organizing map model. The winning neuron is determined based on the preset expected target and the preset evaluation function, and the emotion factor is output based on the winning neuron.

[0090] Among them, the user can define a measurable expected goal for the entire self-organizing map model according to his or her actual needs, and this application does not impose any restrictions.

[0091] Among them, the user can set the specific calculation method of the evaluation function according to his actual needs. This application does not impose any restrictions. The evaluation function is used to characterize the importance of each node (neuron) of the grid structure, and the winning neuron is determined based on this. The winning neuron belongs to the output layer of the self-organizing map model, and the emotion factor is output through the winning neuron.

[0092] In a specific embodiment, the training process of the self-organizing map model includes:

[0093] Obtain target vector samples and emotion factor samples; map the target vector to a multidimensional grid structure through a self-organizing map model, determine the winning neuron based on a preset expected target and a preset evaluation function, and output the emotion factor based on the winning neuron; compare the consistency of the predicted emotion factor and the emotion factor sample to obtain a comparison result; based on the comparison result, adjust the model parameters of the self-organizing map model until the comparison result continues to remain within a second preset range, and determine that the self-organizing map model training is completed.

[0094] Specifically, the self-organizing map model is created based on the Self-Organizing Maps (SOM), which is an unsupervised learning algorithm widely used in deep learning.

[0095] Among them, the target vector is a two-dimensional vector, that is, the data is stored in a two-dimensional array.

[0096] Specifically, a class including weights for initializing model parameters of the self-organizing map model, a method for training the model, and a method for updating the weights is created in advance, and the model is trained by executing the created class.

[0097] Specifically, the target vector sample and the emotion factor sample are merged to obtain a merged two-dimensional vector sample. The two-dimensional vector sample is input into the input layer of the self-organizing map model, and the weights (i.e., model parameters) of the network weight layer are updated by iterative training using the training model method, and the predicted emotion factor is output through the output layer. After multiple iterative operations, the final self-organizing map model is obtained.

[0098] Among them, through Figure 2 The model structure of the self-organizing map model is briefly illustrated. The model structure includes: an input layer 201, a network weight layer 202 and an output layer 203.

[0099] In a specific embodiment, the target vector and the emotion factor are input into a pre-trained navigation decision model, and before the navigation instruction output by the navigation decision model is obtained, each parameter in the emotion factor is normalized to obtain the normalized emotion factor.

[0100] Specifically, after the emotion factor corresponding to the target vector is obtained based on the self-organizing map model, each parameter of the emotion factor is normalized so that the value range of each parameter is between 0 and 1.

[0101] For example, S = [α, β, γ, δ, ε] = [0.7, 0.8, 0.5, 0.1, 0.6]. From this example, we can see that the values ​​of fear and anger are relatively high, and the value of happiness is the lowest. This indicates that the difficulty of the current navigation of the unmanned ship is relatively high, and there may be many obstacles.

[0102] For another example, [α, β, γ, δ, ε] = [0.2, 0.1, 0.5, 0.8, 0.2]. From this example, we can see that the values ​​of fear, anger, sadness, etc. are relatively low, and the value of happiness is the highest. This means that the difficulty of the current navigation of the unmanned ship is relatively low, there may be fewer obstacles, and it can quickly reach the next navigation node.

[0103] In a specific embodiment, the training process of the navigation decision model includes:

[0104] Obtain target vector samples, emotion factor samples and navigation instruction samples; input the target vector samples and emotion factor samples into the navigation decision model to obtain the predicted navigation instruction output by the navigation decision model; obtain the predicted reward based on the predicted navigation instruction, the navigation instruction samples and the preset reward function; based on the predicted reward, adjust the model parameters of the navigation decision model until the predicted reward continues to be maintained within the first preset range, and determine that the navigation decision model training is completed.

[0105] Specifically, the prediction result of the navigation decision model is evaluated using a reward function to assign a reward value (i.e., prediction reward) to it. That is, when the similarity between the predicted navigation instruction and the navigation instruction sample is greater than or equal to the preset similarity, a reward is given for this prediction; when the similarity between the predicted navigation instruction and the navigation instruction sample is less than the preset similarity, the reward is zero.

[0106] The present application utilizes a reward function to guide the navigation decision model to learn desired behaviors during the navigation decision model training process, and can effectively avoid undesirable behaviors, so that the navigation decision model can converge quickly during training.

[0107] Next, through Figure 3 The implementation of this application is described in detail:

[0108] Step 301, obtaining the status information of the unmanned ship and the environmental information of the environment where the unmanned ship is located.

[0109] Step 302: vectorize the state information and environment information of the unmanned ship to obtain a target vector.

[0110] Step 303: input the target vector into the self-organizing map model to obtain the emotion factor output by the self-organizing map model.

[0111] Step 304: input the target vector and the emotion factor into the navigation decision model to obtain the navigation instruction output by the navigation decision model.

[0112] Step 305, based on the navigation instruction, control the unmanned ship to navigate.

[0113] This application constructs and introduces a self-organizing mapping model to obtain emotional factors corresponding to the state information and environmental information of the unmanned ship, and uses emotional factors to guide the training and decision-making process of the enhanced navigation decision model. Using emotional factors to guide the training of the model during the training phase can improve the utilization rate and exploration efficiency of sample data, accelerate the convergence speed of the decision algorithm, and enhance the environmental adaptability of the navigation decision model in complex scenarios.

[0114] Furthermore, the navigation decision model obtained based on the above training can quickly generate a navigation strategy that matches the state information of the unmanned ship and the environmental information of the environment in which the unmanned ship is located, with high flexibility and adaptability.

[0115] The embodiment of the present application also provides an unmanned ship navigation control device based on emotion factors. The specific implementation of the device can refer to the description of the unmanned ship navigation control method based on emotion factors, and the repeated parts will not be repeated. Figure 4 As shown, the device comprises:

[0116] An acquisition module 401 is used to acquire the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and to construct a target vector based on the state information of the unmanned ship and the environmental information;

[0117] The processing module 402 is used to input the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotion factor sample;

[0118] The decision module 403 is used to input the target vector and the emotion factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample;

[0119] The control module 404 is used to control the navigation of the unmanned ship based on the navigation instruction.

[0120] In a specific embodiment, the acquisition module 401 is also used to obtain the route to be navigated corresponding to the unmanned ship, wherein the route to be navigated includes: at least two navigation nodes and route indication information between any two navigation nodes; determining the current navigation node corresponding to the unmanned ship based on the unmanned ship status information; determining the next navigation node corresponding to the unmanned ship based on the route to be navigated and the current navigation node, and the target route indication information between the current navigation node and the next navigation node; the control module 403 is specifically used to control the navigation of the unmanned ship based on the navigation instructions and the target route indication information.

[0121] In a specific embodiment, the navigation instruction includes: the navigation speed of the unmanned ship and the navigation direction of the unmanned ship; the target route indication information includes: the target route indication direction; the control module 403 is specifically used to determine whether the unmanned ship deviates based on the navigation direction and the target route indication direction; when it is determined that the unmanned ship does not deviate, the navigation of the unmanned ship is controlled based on the navigation speed and the navigation direction; when it is determined that the unmanned ship deviates, the navigation of the unmanned ship is controlled based on the navigation speed and the target route indication direction.

[0122] In a specific embodiment, the neural network model includes: a self-organizing map model; a processing module 402, which is specifically used to input a target vector into the self-organizing map model, map the target vector to a multidimensional grid structure through the self-organizing map model, determine a winning neuron based on a preset expected goal and a preset evaluation function, and output an emotion factor based on the winning neuron.

[0123] In a specific embodiment, the decision module 403 is further configured to perform normalization processing on each parameter in the emotion factor to obtain the normalized emotion factor.

[0124] In a specific embodiment, the device also includes a training module for obtaining target vector samples, emotion factor samples and navigation instruction samples; inputting the target vector samples and emotion factor samples into a navigation decision model to obtain predicted navigation instructions output by the navigation decision model; obtaining predicted rewards based on the predicted navigation instructions, the navigation instruction samples and a preset reward function; based on the predicted rewards, adjusting the model parameters of the navigation decision model until the predicted rewards continue to remain within a first preset range, and determining that the training of the navigation decision model is completed.

[0125] In a specific embodiment, the training module is also used to obtain target vector samples and emotion factor samples; input the target vector samples into the self-organizing map model, map the target vector samples to a multidimensional grid structure through the self-organizing map model, determine the winning neuron based on a preset expected goal and a preset evaluation function, and output a predicted emotion factor based on the winning neuron; compare the consistency of the predicted emotion factor and the emotion factor sample to obtain a comparison result; based on the comparison result, adjust the model parameters of the self-organizing map model until the comparison result continues to remain within a second preset range, and determine that the training of the self-organizing map model is completed.

[0126] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502 and the memory 503 complete mutual communication through the communication bus 504. The processor 501 may call the logic instructions in the memory 503 to execute the unmanned ship navigation control method based on the emotion factor, the method comprising: obtaining the unmanned ship state information and the environmental information of the environment where the unmanned ship is located, and constructing a target vector based on the unmanned ship state information and the environmental information; inputting the target vector into a pre-trained neural network model to obtain the emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotion factor sample; inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample; and controlling the navigation of the unmanned ship based on the navigation instruction.

[0127] In addition, the logic instructions in the above-mentioned memory 503 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0128] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the unmanned ship navigation control method based on emotional factors provided by the above methods, and the method includes: obtaining the unmanned ship status information and the environmental information of the unmanned ship environment, and constructing a target vector based on the unmanned ship status information and the environmental information; inputting the target vector into a pre-trained neural network model to obtain the emotional factor output by the neural network model, wherein the emotional factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotional factor sample; inputting the target vector and the emotional factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotional factor sample and the navigation instruction sample; based on the navigation instruction, controlling the navigation of the unmanned ship.

[0129] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the unmanned ship navigation control method based on emotion factors provided in the above-mentioned embodiments, the method comprising: obtaining the unmanned ship status information and the environmental information of the environment in which the unmanned ship is located, and constructing a target vector based on the unmanned ship status information and the environmental information; inputting the target vector into a pre-trained neural network model to obtain the emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotion factor sample; inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample; and controlling the navigation of the unmanned ship based on the navigation instruction.

[0130] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0131] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0132] Finally, it should be noted that the above is only the preferred implementation of the present application, and the present application is not limited to the above embodiments. It is understood that other improvements and changes directly derived or associated by those skilled in the art without departing from the spirit and concept of the present application should be considered to be included in the scope of protection of the present application.

Claims

1. A navigation control method for an unmanned ship based on emotional factors, characterized in that: The method comprises: Acquire the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and construct a target vector based on the state information of the unmanned ship and the environmental information; Inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on target vector samples and emotion factor samples; Inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain a navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample; Based on the navigation instruction, the unmanned ship is controlled to navigate.

2. The unmanned ship navigation control method based on emotional factors according to claim 1 is characterized in that: Before obtaining the unmanned ship status information and the environment information of the unmanned ship, the method further includes: Acquire a route to be navigated corresponding to the unmanned ship, wherein the route to be navigated includes: at least two navigation nodes and route indication information between any two navigation nodes; After obtaining the environmental information of the unmanned ship's environment, it also includes: Based on the unmanned ship state information, determining a current navigation node corresponding to the unmanned ship; Based on the to-be-navigated route and the current navigation node, determining the next navigation node corresponding to the unmanned ship, and target route indication information between the current navigation node and the next navigation node; The step of controlling the unmanned ship to sail based on the navigation instruction includes: Based on the navigation instruction and the target route indication information, the unmanned ship is controlled to navigate.

3. The unmanned ship navigation control method based on emotional factors according to claim 2 is characterized in that: The navigation instruction includes: the navigation speed of the unmanned ship and the navigation direction of the unmanned ship; The target route indication information includes: target route indication direction; The step of controlling the navigation of the unmanned ship based on the navigation instruction and the target route indication information includes: Based on the navigation direction and the target route indication direction, determining whether the unmanned ship deviates; When it is determined that the unmanned ship does not deviate, controlling the navigation of the unmanned ship based on the navigation speed and the navigation direction; When it is determined that the unmanned ship deviates, the navigation of the unmanned ship is controlled based on the navigation speed and the target route indication direction.

4. The unmanned ship navigation control method based on emotion factors according to any one of claims 1 to 3, characterized in that: The neural network model includes: a self-organizing map model; The step of inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model includes: The target vector is input into the self-organizing map model, the target vector is mapped into a multidimensional grid structure through the self-organizing map model, a winning neuron is determined based on a preset expected target and a preset evaluation function, and the emotion factor is output based on the winning neuron.

5. The unmanned ship navigation control method based on emotion factors according to any one of claims 1 to 3, characterized in that: Before inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain the navigation instruction output by the navigation decision model, the method further includes: Each parameter in the emotion factor is normalized to obtain a normalized emotion factor.

6. The unmanned ship navigation control method based on emotion factors according to any one of claims 1 to 3, characterized in that: The training process of the navigation decision model includes: Acquire the target vector sample, the emotion factor sample and the navigation instruction sample; Inputting the target vector sample and the emotion factor sample into the navigation decision model to obtain a predicted navigation instruction output by the navigation decision model; Obtaining a predicted reward based on the predicted navigation instruction, the navigation instruction sample and a preset reward function; Based on the predicted reward, the model parameters of the navigation decision model are adjusted until the predicted reward is continuously maintained within a first preset range, and it is determined that the training of the navigation decision model is completed.

7. The unmanned ship navigation control method based on emotional factors according to claim 4 is characterized in that: The training process of the self-organizing map model includes: Acquire the target vector sample and the emotion factor sample; Inputting the target vector sample into the self-organizing map model, mapping the target vector sample into a multidimensional grid structure through the self-organizing map model, determining a winning neuron based on a preset desired goal and a preset evaluation function, and outputting a predicted emotion factor based on the winning neuron; Comparing the consistency of the predicted emotion factor and the emotion factor sample to obtain a comparison result; Based on the comparison result, the model parameters of the self-organizing map model are adjusted until the comparison result is continuously maintained within a second preset range, and it is determined that the training of the self-organizing map model is completed.

8. An unmanned ship navigation control device based on emotional factors, characterized in that: The device comprises: An acquisition module, used to acquire the state information of the unmanned ship and the environmental information of the environment where the unmanned ship is located, and to construct a target vector based on the state information of the unmanned ship and the environmental information; A processing module, used for inputting the target vector into a pre-trained neural network model to obtain an emotion factor output by the neural network model, wherein the emotion factor is used to characterize the difficulty of the current navigation of the unmanned ship, and the neural network model is trained based on the target vector sample and the emotion factor sample; A decision module, used for inputting the target vector and the emotion factor into a pre-trained navigation decision model to obtain a navigation instruction output by the navigation decision model, wherein the navigation decision model is trained based on the target vector sample, the emotion factor sample and the navigation instruction sample; A control module is used to control the navigation of the unmanned ship based on the navigation instruction.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the unmanned ship navigation control method based on emotion factors as described in any one of claims 1 to 7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the unmanned ship navigation control method based on emotion factors as described in any one of claims 1 to 7 are implemented.

Images