Environment adjusting method and device for ship control cabin and electronic equipment

Abstract

The invention discloses an environment adjusting method and device for a ship control cabin and electronic equipment, and belongs to the technical field of ship control. The method comprises the following steps: acquiring activity data of an operator in a ship control cabin and environment information of the control cabin; extracting working state information of an operator from the activity data; the working state information and the environment information are input into a parameter prediction model, environment adjustment parameters output by the parameter prediction model are obtained, the parameter prediction model is obtained through a plurality of training samples, and the training samples comprise sample working state information and sample environment information of a sample operator; the environment adjusting parameter label is located in a sample control cabin where a sample operator is located; and based on the environment adjustment parameters, calling a mirror image model of an operator to perform analogue simulation, and setting the control cabin through the environment adjustment parameters under the condition that a simulation result of the mirror image model reaches a preset simulation condition. The method can fully consider the requirements of different operators on the environment.

Description

Technical Field

[0001] The present application belongs to the field of ship control technology, and in particular relates to an environmental adjustment method, device and electronic equipment for a ship control cabin. Background Art

[0002] Ship control cabins are typically designed as semi-enclosed or controlled enclosed environments. Specialized ship types may also feature highly enclosed or fully enclosed environments. During missions, operators are required to work in these confined and confined spaces for extended periods of time. Therefore, environmental comfort is crucial to the crew's physical and mental health, work efficiency, and emergency response capabilities. Because control cabins on specialized ship types lack direct access to the external environment, regulating natural environmental factors such as air circulation, temperature, and humidity becomes particularly important. The airtight nature of the control cabin means that fluctuations in oxygen and carbon dioxide concentrations, temperature, and humidity can directly impact the crew's physical and mental condition, mental state, and work efficiency. Any changes in these environmental parameters can cause physical discomfort, such as fatigue, dizziness, and hypothermia, and can even impair their judgment and emergency response capabilities. In severe cases, this can threaten mission success and the crew's lives.

[0003] At present, in order to ensure the health and safety of crew members, sensors are generally used to monitor the environmental parameters and physiological status of operators in the control cabin, and automatic control algorithms are introduced to adaptively adjust the temperature and humidity to ensure that environmental factors remain within an appropriate range and avoid health problems caused by environmental discomfort.

[0004] However, there are significant individual differences between different operators, and the above adjustment scheme relies on a standardized model and cannot fully consider the environmental needs of different operators. Summary of the Invention

[0005] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application proposes an environmental adjustment method, device and electronic equipment for a ship control cabin, which can fully consider the environmental requirements of different operators.

[0006] In a first aspect, the present application provides a method for adjusting the environment of a ship control cabin, the method comprising:

[0007] Acquiring activity data of operators in a ship control cabin and environmental information of the control cabin;

[0008] extracting the operator's work status information from the activity data;

[0009] Inputting the working state information and the environmental information into a parameter prediction model to obtain the environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained through a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and an environmental adjustment parameter label in a sample control cabin where the sample operator is located;

[0010] Based on the environmental adjustment parameters, the operator's mirror model is called for simulation. When the simulation result of the mirror model reaches the preset simulation conditions, the control cabin is set according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics and behavioral characteristics of the operator.

[0011] According to one embodiment of the present application, extracting the operator's work status information from the activity data includes:

[0012] aligning the activity data;

[0013] Mapping the aligned activity data to a target graph structure, where nodes in the target graph structure are used to represent data points in the activity data, and edges between nodes in the target graph structure are used to represent adjacent relationships and similarities between data points;

[0014] The target graph structure is feature extracted through a graph attention network to obtain the working status information.

[0015] According to one embodiment of the present application, the scale alignment of the activity data includes:

[0016] Call the inference engine to reason about the active data and determine the rule category and processing agenda;

[0017] Based on the rule category, calling corresponding processing rules from a rule library to match each category of data in the activity data;

[0018] Invoking an execution engine to process corresponding activity data according to the processing rules based on the processing agenda, and determining sensitive data and non-sensitive data in the activity data;

[0019] extracting risk assessment features from the sensitive data based on the relevance between the sensitive data and the operator's working status;

[0020] The covariance matrices of the non-sensitive data and the risk assessment features are scale-aligned.

[0021] According to one embodiment of the present application, the simulation of the mirror model is performed based on a state transition model, which includes a domain-invariant feature extractor, a cross-modal adversarial fusion module, and an environmental parameter prediction engine connected in sequence;

[0022] The step of calling the operator's mirror model to perform simulation based on the environmental adjustment parameters, and setting the control cabin according to the environmental adjustment parameters when the simulation result of the mirror model meets a preset simulation condition, includes:

[0023] Performing multi-scale feature extraction on the activity data and the environmental information using the domain-invariant feature extractor to obtain a feature vector;

[0024] Performing adversarial training and fusing the feature vectors through the cross-modal adversarial fusion module to generate fused features;

[0025] Through the environmental parameter prediction engine, the fusion features are forward migrated to the mirror model, and the mirror model is simulated based on the environmental adjustment parameters. When it is detected that the simulation results deviate from the preset simulation conditions, reverse migration is triggered, and the parameter prediction model is optimized to obtain new environmental adjustment parameters until the simulation results meet the preset simulation conditions. The control cabin is configured according to the new environmental adjustment parameters.

[0026] According to one embodiment of the present application, the domain-invariant feature extractor includes a ResNet-18 network and a gradient reversal layer connected in sequence;

[0027] The ResNet-18 network is used to perform multi-scale feature extraction on the activity data and environmental information to obtain a feature vector;

[0028] The gradient reversal layer is used to perform bidirectional state migration and feature alignment on the feature vector.

[0029] According to one embodiment of the present application, when it is detected that the simulation result deviates from the preset simulation condition, triggering reverse migration and optimizing the parameter prediction model includes:

[0030] When the simulation result deviates from the preset simulation condition, the incremental memory replay mechanism is triggered to calculate the historical optimal parameter combination corresponding to the activity data and the environmental information by replaying the historical data in the core memory bank and the edge memory bank;

[0031] The parameter prediction model is elastically weighted and updated based on the optimal parameter combination.

[0032] According to one embodiment of the present application, before calling the operator's mirror model to perform simulation based on the environment adjustment parameters, the method further includes:

[0033] Pre-train the state transfer model using source domain data;

[0034] Freezing the domain-invariant feature extractor and the environmental parameter prediction engine, and fine-tuning the parameters of the cross-modal adversarial fusion module using target domain data;

[0035] The domain-invariant feature extractor and the environmental parameter prediction engine are gradually unfrozen, and the parameters of the state transition model are optimized using target domain data.

[0036] According to one embodiment of the present application, obtaining activity data of an operator in a ship control cabin includes:

[0037] acquiring physiological data, psychological data, and behavioral data of an operator in the control cabin;

[0038] calculating a signal quality index for the physiological data, psychological data, and behavioral data, and filtering the physiological data, psychological data, and behavioral data to obtain the activity data when the signal quality index is greater than a preset threshold;

[0039] The control cabin configures the environmental adjustment parameters through an environmental adjustment actuator. After the control cabin is set using the environmental adjustment parameters, the method further includes:

[0040] Retrieving the operator's activity data;

[0041] The environmental adjustment actuator is feedback-monitored through an actuator feedback verification loop, and when it is determined that the error between the new activity data and the simulation result is greater than an error threshold, the control cabin parameters are adjusted through the environmental adjustment actuator.

[0042] In a second aspect, the present application provides an environmental conditioning device for a ship control cabin, the device comprising:

[0043] an acquisition module, configured to acquire activity data of an operator in a control cabin of a ship and environmental information of the control cabin;

[0044] a first processing module, configured to extract the operator's working status information from the activity data;

[0045] a second processing module, configured to input the working state information and the environmental information into a parameter prediction model to obtain environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained through a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and an environmental adjustment parameter label in a sample control cabin where the sample operator is located;

[0046] The third processing module is used to call the operator's mirror model for simulation based on the environmental adjustment parameters. When the simulation result of the mirror model meets the preset simulation conditions, the control cabin is set according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics and behavioral characteristics of the operator.

[0047] In a third aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the environmental adjustment method for the ship control cabin as described in the first aspect above is implemented.

[0048] In a fourth aspect, the present application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the environmental adjustment method for the ship control cabin as described in the first aspect above.

[0049] In a fifth aspect, the present application provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the environmental adjustment method for the ship control cabin as described in the first aspect.

[0050] In a sixth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the environmental adjustment method for the ship control cabin as described in the first aspect above.

[0051] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become obvious from the description below, or will be learned through practice of the present application.

[0052] The environmental adjustment method, device, and electronic equipment for a ship control cabin provided in this application have the following beneficial effects compared to the prior art:

[0053] (1) By real-time monitoring of environmental information and operator activity data, the operator's fatigue and distraction can be discovered in time, the environmental adjustment parameters can be verified through the operator's personalized mirror model, the operator's reaction in a specific environment can be simulated, and the environmental adjustment parameters can be adjusted in real time according to their physiological and psychological state. This can fully consider the environmental needs of different operators, ensure that the operator can maintain the best working state after working for a long time, and ensure that the environmental adjustment parameters more accurately meet the needs of each operator, thereby improving the operator's work efficiency and task completion quality, as well as the overall operational efficiency of the ship's control cabin, and further optimizing the control cabin environment. This can not only improve the operator's working state, but also improve the safety and stability of the ship, and prevent mistakes or accidents caused by the operator's inability to adapt to the environment.

[0054] (2) Through bidirectional state migration, cross-modal fusion and dynamic parameter updating, combined with the bidirectional migration mechanism of domain-invariant feature extraction, cross-modal adversarial fusion and environmental parameter prediction engine, the accuracy, adaptability and stability of the mirror model in simulation are improved, thereby optimizing the real-time adjustment of the control cabin configuration. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0056] Figure 1 1 is a flow chart of an environmental adjustment method for a ship control cabin provided in an embodiment of the present application;

[0057] Figure 2 It is a structural diagram of the state transition model provided in an embodiment of the present application;

[0058] Figure 3 This is a schematic structural diagram of an environmental adjustment device for a ship control cabin provided in an embodiment of the present application;

[0059] Figure 4 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described 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 are within the scope of protection of this application.

[0061] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0062] Below, in combination with the accompanying drawings, the environmental adjustment method for the ship control cabin, the environmental adjustment device for the ship control cabin, the electronic device and the readable storage medium provided in the embodiments of the present application are described in detail through specific embodiments and their application scenarios.

[0063] The environmental adjustment method for the ship control cabin may be applied to a terminal, and may be specifically executed by hardware or software in the terminal.

[0064] The terminal includes, but is not limited to, a portable communication device such as a mobile phone or tablet computer having a touch-sensitive surface (e.g., a touch screen display and / or a touch pad). It should also be understood that, in some embodiments, the terminal may not be a portable communication device, but a desktop computer having a touch-sensitive surface (e.g., a touch screen display and / or a touch pad).

[0065] In the following embodiments, a terminal including a display and a touch-sensitive surface is described. However, it should be understood that the terminal may include one or more other physical user interface devices such as a physical keyboard, a mouse, and a joystick.

[0066] The embodiment of the present application provides an environmental adjustment method for a ship control cabin. The execution subject of the environmental adjustment method for a ship control cabin may be an electronic device or a functional module or functional entity in the electronic device that can implement the environmental adjustment method for the ship control cabin. The electronic devices mentioned in the embodiment of the present application include but are not limited to mobile phones, tablet computers, computers, cameras, and wearable devices. The environmental adjustment method for a ship control cabin provided in the embodiment of the present application is described below using an electronic device as an example of the execution subject.

[0067] like Figure 1 As shown, the environmental adjustment method of the ship control cabin includes:

[0068] Step 110: Acquire activity data of operators in the ship control cabin and environmental information of the control cabin;

[0069] Step 120: extracting the operator's work status information from the activity data;

[0070] Step 130: Input the working state information and the environmental information into a parameter prediction model to obtain the environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained by using a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and the environmental adjustment parameter label of the sample control cabin where the sample operator is located;

[0071] Step 140: Based on the environmental adjustment parameters, call the operator's mirror model for simulation. When the simulation result of the mirror model meets the preset simulation conditions, set the control cabin according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics, and behavioral characteristics of the operator.

[0072] It is understandable that the ship's control cabin is a closed working environment, and the operator needs to carry out activities and work in the control cabin for a long time to control the ship. The environmental adjustment parameters in the control cabin need to be adjusted accordingly to improve the operator's status.

[0073] Activity data refers to physiological data collected from operators through sensors, as well as behavioral data and other parameters collected from operators through operation recording systems.

[0074] Physiological data includes the operator's heart rate, body temperature, pupil tracking data, and brain waves. These data reflect the operator's physiological state, such as fatigue and attention. Operational behavior data includes operation frequency, device interaction trajectory, command response delay, etc. Operator behavior data, such as operation frequency, device interaction trajectory, and command response delay, is collected through the operation recording system. This data reflects the operator's behavioral state and work efficiency.

[0075] Environmental information is a vector composed of environmental data such as temperature, humidity, oxygen concentration, carbon dioxide concentration, air pressure, and air circulation conditions in the control cabin. Through environmental data, it is possible to identify whether the environment inside the control cabin is suitable, whether it is too hot, cold, or humid, affecting the operator's physiological and psychological state.

[0076] Work status information is a quantitative indicator extracted from activity data, including physiological, psychological, and work indicators related to the operator's work status. For example, fatigue can be extracted from heart rate variability, attention can be extracted from pupil tracking data, and operational efficiency can be extracted from command response latency.

[0077] Among them, pupil tracking data is recorded by the ultra-wide band (UWB) positioning system to record the operation trajectory, combined with the attention focus captured by the eye tracker.

[0078] The parameter prediction model is built based on a neural network model. During the training process, a large number of training samples are collected. Each sample data includes the sample operator's optimal working state, sample environmental information, and the corresponding environmental adjustment parameter label. The training samples come from different operators and different situations in the ship's control room. The training samples are sequentially input into the neural network model for training. Based on the input working state and environmental information, the neural network model can infer the most appropriate environmental adjustment parameters for each operator.

[0079] Environmental adjustment parameters are a combination of the set values ​​of multiple devices in the control cabin. Through the coordinated control of multiple set values, the temperature, humidity, oxygen content, lighting intensity, air pressure, etc. in the operating cabin are adjusted. For example, the CO2 concentration linkage mechanism of ship ventilation control is adopted to optimize the fresh air circulation efficiency, integrate the temperature and humidity domain control algorithm, and use Fluent software to simulate the impact of airflow distribution on operating comfort, avoid sudden changes in local environmental data, and reduce ineffective energy consumption.

[0080] The mirror model is constructed based on the operator's physiological, psychological and behavioral characteristics. Physiological characteristics include the changing trends of the operator's physiological data such as body temperature and heart rate; psychological characteristics include the operator's emotional state and fatigue level, which are evaluated through long-term sleep monitoring; behavioral characteristics include the operator's activity level, rest pattern, movement pattern and reaction speed, etc. The mirror model also takes into account individual differences among operators, such as age, physique, emotional tendencies and other factors, and can more accurately simulate the reactions of each operator in different environments.

[0081] For example, wearable devices such as smart bracelets and health bracelets collect the operator's physiological data, such as heart rate, blood pressure, blood oxygen saturation, and body temperature, and transmit it to the computing platform in real time. The computing platform efficiently processes and analyzes this data to obtain the operator's physiological, psychological, and behavioral characteristics, and then models the operator. Using virtual mirror technology, the operator's different factors are accurately mapped to the digital twin in the virtual control cabin. Personalized training is then conducted based on physiological, psychological, and behavioral characteristics. The resulting mirror model is used to simulate the operator's state and response to external stimuli to accurately reflect the operator's physiological state and health status. For example, in an emergency, the operator may be more sensitive to changes in temperature and oxygen concentration, while in normal working conditions, the reaction may be more stable.

[0082] In addition, the mirror model integrates a computational fluid dynamics (CFD) simulation module. The CFD simulation module can accurately simulate physical phenomena such as fluid flow and heat transfer, providing authenticity support for physical laws for digital twins. Through multi-physics field simulation, it verifies the impact of environmental adjustment parameters on the operator's human-computer interaction in the control cabin.

[0083] The preset simulation conditions are dynamic thresholds used to judge the effectiveness of environmental adjustment parameters. For example, work efficiency is improved by more than 15%, or fatigue is reduced to a safe range, the temperature is within the operator's comfort range, the humidity is moderate, etc. The operator's preset simulation conditions are adjusted using cyclic adjustment logic combined with an adaptive threshold adjustment mechanism.

[0084] In addition, a 1:1 control cabin model was built in the Unity3D engine, and the mirror model loaded the operator's biomechanical properties in the control cabin model to simulate the environmental adjustment parameters.

[0085] In step 110, multiple sensors are deployed in the ship's control cabin, including temperature sensors, humidity sensors, air quality sensors, air pressure sensors, etc., to collect environmental data to obtain environmental information. Various biosensors such as heart rate monitors, brain wave sensors, pupil tracking devices, etc. are embedded in the console, seats, and helmets to monitor the operator's heart rate, galvanic skin response and other physiological data. The operation recording system collects the operator's behavioral data, and edge computing nodes are used to perform spatiotemporal alignment of multi-source data, and the combination of physiological data and behavioral data is used as activity data.

[0086] In step 120, the operator's work status information is extracted from the sensor data using machine learning or pattern recognition techniques by analyzing the activity data. For example, by analyzing heart rate and movement data, the operator's work intensity can be inferred as light work, moderate work, or heavy work. A wavelet transform is used to extract the entropy of physiological fluctuations from the activity data to determine fatigue level. Fractal mantissa calculation is used to extract the complexity of the operation trajectory from the activity data to obtain attention concentration and operation efficiency. Work intensity, fatigue level, attention concentration, and operation efficiency are then used as work status information.

[0087] In step 130, the extracted working status information and environmental information are input into a parameter prediction model, and the most suitable environmental adjustment parameters for the operator are calculated and output through the trained parameter prediction model.

[0088] In step 140, the operator's activity data and environmental data are input into the mirror model to simulate their response to the environmental conditions corresponding to the environmental adjustment parameters, ensuring that their response meets the preset conditions. For example, when the temperature or humidity in the control cabin is too high, the operator's comfort level may decrease. The mirror model can reflect this change. When the simulation results of the mirror model meet the preset simulation conditions, the environmental adjustment parameters are used to adjust the control cabin environment to optimize the control cabin environment. This includes adjusting at least one of the following: air conditioning, ventilation system, humidification or dehumidification equipment, air purification system, and lighting system. This improves the operator's working state and the work efficiency of the ship's control cabin, reduces operator fatigue and psychological stress, and maintains the operator's high alertness and work enthusiasm.

[0089] In addition, the control cabin is equipped with a feedback mechanism that automatically detects and corrects environmental settings if environmental conditions suddenly change or the operator's physiological response does not meet expectations, ensuring that the environment in the control cabin always meets the needs of the crew.

[0090] For example, when the operator's working status changes, such as entering a high-load state or a rest state, the operator's activity data and the environmental information of the control cabin are reacquired, the parameter prediction model is called to recalculate the environmental adjustment parameters according to the newly input data, and the mirror model is called for simulation to adjust the environment inside the ship cabin.

[0091] To improve the environmental conditioning effect of the ship's control cabin, a temperature and humidity disturbance dataset dynamically controlled in a low-pressure test chamber is used to simulate data disturbances under ship turbulence conditions for adversarial training. This covers extreme conditions such as deep-diving high-pressure environments. An inertial measurement unit (IMU) can also be added, integrated with an attitude compensation algorithm, to consider the impact of ship turbulence acceleration on operations.

[0092] A floating anomaly correction mechanism can also be set up to trigger the parameter lock of the operation cabin when the simulated operation error rate drops by ≥20%.

[0093] According to the environmental adjustment method for the ship control cabin provided in the embodiment of the present application, by real-time monitoring of environmental information and operator activity data, the operator's fatigue and distraction can be discovered in time, the environmental adjustment parameters can be verified through the operator's personalized mirror model, the operator's reaction in a specific environment is simulated, and the environmental adjustment parameters are adjusted in real time according to their physiological and psychological state. This can fully consider the environmental needs of different operators, ensure that the operator can maintain the best working state after working for a long time, and ensure that the environmental adjustment parameters more accurately meet the needs of each operator, thereby improving the operator's work efficiency and task completion quality, as well as the overall operational efficiency of the ship control cabin, and further optimizing the environment of the control cabin, which can not only improve the operator's working state, but also improve the safety and stability of the ship, and prevent mistakes or accidents caused by the operator's inability to adapt to the environment.

[0094] In some embodiments, step 120, extracting the operator's work status information from the activity data, includes:

[0095] aligning the activity data;

[0096] Mapping the aligned activity data to a target graph structure, where nodes in the target graph structure are used to represent data points in the activity data, and edges between nodes in the target graph structure are used to represent adjacent relationships and similarities between data points;

[0097] The target graph structure is feature extracted through a graph attention network to obtain the working status information.

[0098] Among them, the activity data includes physiological data, psychological data and behavioral data. The behavioral data includes electrocardiogram (ECG), brain wave (EEG), heart rate, blood oxygen, body temperature, etc. The psychological data is used to reflect the operator's psychological state, such as mood swings, psychological pressure, anxiety level, fatigue level, etc. The behavioral data includes data related to the operator's behavior, actions, and work task execution, such as mouse clicks, keyboard input, eye movement trajectory, posture changes, operation response time, etc.

[0099] Since the sampling frequencies of physiological data and behavioral data are different, and psychological data is discrete data, the physiological data, psychological data, and behavioral data are aligned to the same time point to achieve temporal alignment.

[0100] The target graph structure is a graphical data structure used to represent the relationship between different data points, including nodes and edges between nodes. Nodes are used to represent data points. Each data point in physiological data, psychological data, and behavioral data can be mapped to a node. Edges and weights are used to represent the adjacent relationship or similarity between nodes. Edges are constructed based on the correlation or distance measurement between physiological data, psychological data, and behavioral data.

[0101] Physiological indicators are used to describe the characteristics of the operator's physiological state and are obtained by extracting key features from physiological data. Psychological indicators are used to describe the characteristics of the operator's psychological state. Work efficiency is an indicator used to evaluate the quality and speed of the operator's task completion. By integrating physiological and behavioral data, the relationship between the operator's work state and work efficiency is quantified.

[0102] In actual implementation, the physiological data, psychological data, and behavioral data are aligned in time scale through interpolation. For time series data, edges between adjacent nodes are established according to the time order. The similarity between nodes is measured using Euclidean distance or cosine similarity, and corresponding weights are assigned to the edges to obtain the target graph structure. The Graph Attention Network (GAT) uses a self-attention mechanism to learn the features of nodes and dynamically adjust the weights between adjacent nodes.

[0103] Based on temporal adjacency and similarity calculations, or task-based behavioral similarity, the edges in the target graph structure are represented by an adjacency matrix. Each element of the adjacency matrix represents the similarity or relationship strength between nodes.

[0104] In the graph attention network, each node dynamically calculates the weights of its neighboring nodes through the self-attention mechanism based on the characteristics of its neighboring nodes, automatically identifies the neighboring nodes that have a greater impact on the current node, and thus effectively extracts the node features; through multiple graph convolutional layers (Graph ConvolutionalLayers) to pass information layer by layer, high-order node features are extracted, such as frequency domain features extracted from EEG signals, or time features extracted from behavioral data.

[0105] The extracted features are used to evaluate the operator's working status, and indicators such as physiological load, psychological state, and work efficiency are output as working status information.

[0106] In this embodiment, by constructing a target graph structure, the operator's work status information can be effectively extracted from multi-dimensional data, providing a basis for decision support, health management and work optimization.

[0107] In some embodiments, aligning the activity data by scale includes:

[0108] Call the inference engine to reason about the active data and determine the rule category and processing agenda;

[0109] Based on the rule category, calling corresponding processing rules from a rule library to match each category of data in the activity data;

[0110] Invoking an execution engine to process corresponding activity data according to the processing rules based on the processing agenda, and determining sensitive data and non-sensitive data in the activity data;

[0111] extracting risk assessment features from the sensitive data based on the relevance between the sensitive data and the operator's working status;

[0112] The covariance matrices of the non-sensitive data and the risk assessment features are scale-aligned.

[0113] It can be understood that the reasoning engine is used to reason and analyze activity data, determine the rule category corresponding to each type of activity data, and generate a processing agenda for the activity data in the subsequent classification order. The reasoning engine can be a rule-based reasoning engine or a machine learning-based reasoning engine.

[0114] Rule categories are used to characterize the operator's behavior patterns, changes in environmental parameters, etc. corresponding to each type of data, and the processing agenda is the processing steps and processes of active data.

[0115] The rule base contains multiple processing rules and is a structured set of rules that supports category-based matching and the invocation of specific rules. Processing rules can be categorized and defined based on operator behavior, environmental parameters, and other factors, identifying sensitive data that has a significant impact on the operator's work status and other non-sensitive data.

[0116] For example, if the data indicates that the operator is in an abnormal working state such as unauthorized operation or long-term overtime, it is determined to be sensitive data.

[0117] Risk assessment characteristics are used for subsequent risk analysis and decision-making.

[0118] In actual execution, the activity data is input into the inference engine, which infers the activity data based on pre-defined rules or trained models to determine the rule category and processing agenda.

[0119] Based on the determined rule category, the corresponding processing rules are called from the rule library to match each type of data in the activity data. For example, if the rule category is a specific operator operation behavior, then the rules related to the operation behavior are called to match the corresponding data in the activity data.

[0120] The execution engine is called to process the corresponding activity data according to the data processing steps and processing logic built into the processing agenda through the matched processing rules. During the processing, sensitive data and non-sensitive data are separated.

[0121] Based on the correlation between sensitive data and the operator's work status, highly relevant risk assessment features are extracted from the sensitive data. For example, if the sensitive data is the operator's physiological data, features such as heart rate and blood pressure can be extracted to assess whether the operator's work status is normal.

[0122] Non-sensitive data and risk assessment features are standardized, and the covariance matrix is ​​calculated to measure the correlation and degree of change between non-sensitive data and risk assessment features, so as to achieve scale alignment. This allows different data to be compared and analyzed on the same scale, providing a basis for subsequent data processing and decision-making.

[0123] In this example, by reasoning, classifying, and scalar-aligning activity data, the secure processing of sensitive data and the proper storage of non-sensitive data are ensured. Furthermore, the alignment of risk assessment features and covariance matrices provides a foundation for subsequent analysis and decision-making.

[0124] In some embodiments, the simulation of the mirror model is based on a state transition model, such as Figure 2 As shown, the state transition model includes a domain-invariant feature extractor, a cross-modal adversarial fusion module, and an environment parameter prediction engine connected in sequence;

[0125] Step 140, calling the operator's mirror model to perform simulation based on the environmental adjustment parameters, and setting the control cabin according to the environmental adjustment parameters when the simulation result of the mirror model meets the preset simulation conditions, includes:

[0126] Performing multi-scale feature extraction on the activity data and the environmental information using the domain-invariant feature extractor to obtain a feature vector;

[0127] Performing adversarial training and fusing the feature vectors through the cross-modal adversarial fusion module to generate fused features;

[0128] Through the environmental parameter prediction engine, the fusion features are forward migrated to the mirror model, and the mirror model is simulated based on the environmental adjustment parameters. When it is detected that the simulation results deviate from the preset simulation conditions, reverse migration is triggered, and the parameter prediction model is optimized to obtain new environmental adjustment parameters until the simulation results meet the preset simulation conditions. The control cabin is configured according to the new environmental adjustment parameters.

[0129] Among them, bidirectional state migration includes forward migration and reverse migration.

[0130] It can be understood that the state transition model is a physiological-psychological state transition model (PPSTM), which is first trained on a large-scale cross-operator dataset and then fine-tuned for each operator. The resulting state transition model can reflect the individual differences of the operators.

[0131] In some embodiments, the domain-invariant feature extractor includes a ResNet-18 network and a gradient reversal layer connected in sequence;

[0132] The ResNet-18 network is used to perform multi-scale feature extraction on the activity data and environmental information to obtain a feature vector;

[0133] The gradient reversal layer is used to perform bidirectional state migration and feature alignment on the feature vector.

[0134] The domain-invariant feature extractor is constructed based on the Adversarial Discriminative Domain Adaptation (ADDA) framework by adding a gradient reversal layer (GRL) to the ResNet-18 output layer.

[0135] The ResNet-18 network extracts multi-scale features from the input data through multi-layer convolution and residual block structures to obtain feature vectors of multiple scales. While maintaining high performance, it solves the problems of gradient vanishing and gradient exploding in deep networks. Through residual learning and skip connections, the network can learn feature representations more effectively.

[0136] In domain adaptation tasks, the Gradient Reversal Layer (GRL) reverses the sign of the gradient during backpropagation, enabling the domain-invariant feature extractor to learn domain-invariant features. The extracted feature vectors are more consistent between the source and target domains. This mechanism makes it difficult for the domain classifier to distinguish between the features of the source and target domains, thereby improving the model's generalization ability across different domains.

[0137] In actual implementation, the domain-invariant feature extractor utilizes a ResNet-18 network to extract multi-scale features, eliminating data distribution differences between the laboratory (source domain) and the ship's control cabin (target domain) and ensuring feature alignment. For example, temporal and spatial features are extracted for operator physiological data and environmental information, respectively, and then the Maximum Mean Discrepancy (MMD) loss is used to constrain feature distribution consistency.

[0138] Adversarial training with gradient reversal layers (GRLs) forces feature vectors from different modalities to align in the latent space, and dynamically assigns weights through an attention mechanism. In the cross-modal adversarial fusion module, adversarial training effectively mitigates the modality gap, and the strong semantic correlation between feature vectors in the modalities improves fusion. For example, adversarial fusion is performed on EEG spectrum features and the temporal changes in CO2 concentration in the cabin.

[0139] The environmental parameter prediction engine performs forward migration based on the fused features, using a mirror model to predict the operator's state after environmental parameter adjustments. In the reverse migration, the parameter update amount of the parameter prediction model is inferred based on the deviation of the simulation results, forming a closed-loop optimization. For example, if the temperature rise in the operating cabin causes the operation delay rate to exceed a threshold (such as 5%), the reverse migration calculates the optimal temperature control strategy.

[0140] In the forward transfer, activity data and environmental information are mapped into changes in the operator's physiological and psychological states by fusion features, and the causal relationship between parameters and states is modeled through the state transition matrix.

[0141] In reverse migration, counterfactual reasoning is used to trace back the environmental causes that lead to simulation deviations, and the weights of the parameter prediction model are updated. For example, the environmental cause is that the CO2 concentration exceeds the standard, which leads to decision-making errors.

[0142] In this embodiment, through bidirectional state migration, cross-modal fusion and dynamic parameter update, combined with the bidirectional migration mechanism of domain-invariant feature extraction, cross-modal adversarial fusion and environmental parameter prediction engine, the accuracy, adaptability and stability of the mirror model in simulation are improved, thereby optimizing the real-time adjustment of the control cabin configuration.

[0143] In some embodiments, the feature alignment loss function of the state transition model includes an adversarial alignment loss function, a statistical alignment loss function, and a semantic preservation loss function;

[0144] The adversarial alignment loss function L adv for:

[0145]

[0146] Among them, x s and x t Represent the input data of the source domain and the input data of the target domain respectively, D is the domain discriminator, and f is the feature extractor;

[0147] The statistical alignment loss function L stat for:

[0148]

[0149] Where MMD is the maximum mean difference, is the feature mapping function, C s and C t are the covariance matrices of the source and target domains, λ1 and λ2 are weight coefficients, and CORAL is covariance alignment;

[0150] The semantic preservation loss function L task for:

[0151]

[0152] Among them, CE is the cross entropy loss, which is used for the source domain classification task; KL is the Kullback-Leibler divergence, which is used for the target domain knowledge distillation, y s and are the true label and predicted label of the source domain, p t and p teacher are the output probability distributions of the target domain model and the teacher model, respectively.

[0153] In practice, the domain-invariant feature extractor extracts features from both the source and target domains. The feature alignment loss function matches the source and target domain features using alignment methods corresponding to the adversarial alignment loss, statistical alignment loss, and semantic preservation loss, ensuring that the extracted features remain consistent across cross-domain tasks. This helps the feature extractor learn feature representations that are valid for both the source and target domains.

[0154] The cross-modal adversarial fusion module further optimizes feature fusion through adversarial training methods, so that the features of the source domain and the target domain can be effectively integrated to facilitate subsequent environmental parameter prediction. The feature alignment loss function combines adversarial alignment loss, statistical alignment loss, and semantic preservation loss to ensure that cross-modal features can remain consistent to a certain extent, further promoting the adversarial fusion of the source domain and the target domain.

[0155] The environmental parameter prediction engine uses the aligned features to make the final task prediction. The aligned features ensure that the target domain data can also be effectively predicted in the environmental parameter prediction engine.

[0156] Among them, the adversarial alignment loss function L adv By maximizing the probability of correct classification of source domain features and incorrect classification of target domain features by the domain discriminator D, the feature extractor is forced to generate domain-invariant features. A gradient reversal layer (GRL) is introduced to dynamically confuse the judgment of the domain discriminator D.

[0157] Statistic alignment loss function L statIt includes Maximum Mean Discrepancy (MMD) and Correlation Alignment (CORAL). MMD calculates the distribution distance in the reproducing kernel Hilbert space, and CORAL aligns the second-order statistics of the covariance matrix to align the first-order and second-order statistics of the source and target domain features.

[0158] Semantic preservation loss function L task Jointly optimize the source domain classification task and target domain knowledge distillation to ensure that the features extracted by the feature extractor can retain the semantic information of the data and the semantic consistency of the features.

[0159] The above three loss functions are jointly optimized to train the domain-invariant feature extractor. By minimizing the total loss function, the feature extractor can extract domain-invariant and semantically consistent features.

[0160] In this embodiment, the role of the feature alignment loss function is to optimize the domain-invariant feature extractor to ensure that the features of the source domain and the target domain can be aligned after passing through the feature extractor. At the same time, the feature alignment is further optimized in the cross-modal adversarial fusion module, and finally effective and aligned features are provided for the environmental parameter prediction engine, which can achieve better feature representation alignment in cross-domain migration and improve the generalization ability of the state migration model between different domains.

[0161] In some embodiments, when it is detected that the simulation result deviates from the preset simulation condition, triggering reverse migration and optimizing the parameter prediction model includes:

[0162] When the simulation result deviates from the preset simulation condition, the incremental memory replay mechanism is triggered to calculate the historical optimal parameter combination corresponding to the activity data and the environmental information by replaying the historical data in the core memory bank and the edge memory bank;

[0163] The parameter prediction model is elastically weighted and updated based on the optimal parameter combination.

[0164] Among them, the parameter prediction model is updated through a time-series enhanced incremental learning algorithm, and a dual memory playback mechanism is designed, including a core memory library and an edge memory library.

[0165] In the core memory library, key samples are selected based on influence score:

[0166]

[0167] Among them, IS(x i ) is x iInfluence score of i is the i-th input sample; N is the total number of samples in the data set; Indicates that the input sample x i Calculate the gradient of the loss function with respect to the model parameters θ, is the update direction of the model parameters, L(x i ) is the input sample x i The loss function on ; For x j The associated gradient represents the loss function L(x j ) with respect to the model parameters θ; To calculate x i and x j The square of the gradient difference between .

[0168] The core memory library retrieves the most representative samples and uses the information from these samples to adjust the parameters and predict the parameters of the entire model. The core memory library selects key samples through influence scoring and stores the 200 most representative samples.

[0169] In building the edge memory library, the edge memory library stores the data of the last 24 hours and is dynamically updated through a sliding window mechanism. If the Kullback-Leibler divergence of the new data is greater than the threshold, the old data will be eliminated and the model will be adjusted based on these real-time data.

[0170] By combining the advantages of dynamic memory management and edge computing, a balance between real-time response and continuous learning can be achieved in resource-constrained environments.

[0171] In actual execution, when the simulation results deviate from the preset conditions, the state transition model triggers the incremental memory replay mechanism, and calculates the optimal parameter combination through the historical data stored in the core memory library and the edge memory library. The optimal parameter combination is used to represent the configuration in which the parameter prediction model achieves better performance under the conditions of the same activity data and environmental information.

[0172] In time series enhanced incremental learning algorithms, data augmentation techniques are used to enhance the performance of time series data, taking into account the temporal dependencies of historical data. Methods such as data smoothing, time series interpolation, and time windows can be used to enhance the model's ability to process time series data.

[0173] During each incremental learning step, the most important part of reinforcement learning is to replay samples from the memory bank. The model can be updated at each moment by combining memory replay with current data.

[0174] The loss function calculation needs to consider not only the current prediction error but also the impact of historical data playback. The loss function is optimized by combining the playback data with the current model prediction results.

[0175] During the continuous optimization process, the Elastic Weight Consolidation (EWC) algorithm is used to avoid catastrophic forgetting and maintain the stability of historical learning.

[0176] Elastic Weight Curing (EWC) loss function L total for:

[0177]

[0178] Among them, L new is the current loss function, is the importance weight of the historical parameters updated at the kth time, θ k is the current parameter of the kth update, is the historical optimal parameter for the kth update, and λ is a hyperparameter.

[0179] EWC loss function L total Used to punish large changes in parameters, importance weights The Fisher information matrix is ​​calculated jointly with the time decay factor. The Fisher information matrix reflects the influence of each parameter on the model output, while the time decay factor takes into account that parameters that have not been updated for a long time need more penalties.

[0180] By reversing the parameter update process triggered by the migration, the optimization goal is to minimize the bias and enable the model to adapt to new input data.

[0181] This example combines the advantages of dynamic memory management and edge computing to achieve a balance between real-time response and continuous learning in resource-constrained environments. When processing the edge memory, a sliding window mechanism and active forgetting strategy are used to maintain the system's real-time performance while leveraging edge computing resources to quickly respond and update the model.

[0182] In some embodiments, before calling the operator's mirror model to perform simulation based on the environmental adjustment parameters, the method further includes:

[0183] Pre-train the state transfer model using source domain data;

[0184] Freezing the domain-invariant feature extractor and the environmental parameter prediction engine, and fine-tuning the parameters of the cross-modal adversarial fusion module using target domain data;

[0185] The domain-invariant feature extractor and the environmental parameter prediction engine are gradually unfrozen, and the parameters of the state transition model are optimized using target domain data.

[0186] In actual implementation, the state transition model is pre-trained, the domain-invariant feature extractor extracts common features of the source domain data, the cross-modal adversarial fusion module learns how to find shared features between different modalities and enhances the fusion effect between different modalities through adversarial training, and the environmental parameter prediction engine uses the source domain data to preliminarily learn how to predict environmental states and related parameters.

[0187] Freeze the domain-invariant feature extractor and environmental parameter prediction engine, and focus on optimizing the cross-modal adversarial fusion module to better fuse features from different modalities in the target domain. This ensures that the domain-invariant feature extractor and environmental parameter prediction engine of the state transition model are not over-tuned, focusing solely on improving the performance of the cross-modal adversarial fusion module.

[0188] The domain-invariant feature extractor and environmental parameter prediction engine are unfrozen from the bottom layer to the top layer. First, the low-level feature extraction part of the network is unfrozen, and then the high-level structure is gradually unfrozen. The state transition model can gradually adapt to new patterns and feature changes that may exist in the target domain data. During the unfreezing process, the domain-invariant feature extractor, cross-modal adversarial fusion module and environmental parameter prediction engine are gradually updated through training with target domain data to optimize the parameters and overall performance of the entire state transition model.

[0189] In addition, the parameters of each module in the state transition model can be jointly updated through the target domain data, so that the state transition model can better adapt to the task requirements of the target domain.

[0190] In this embodiment, by training the state migration model on the source domain dataset, learning common features, freezing the domain invariant feature extractor and the environmental parameter prediction engine, and only fine-tuning the cross-modal adversarial fusion module, adversarial fusion features between different modalities are learned through adversarial training, and effective common features can be found between multiple modalities, promoting the state migration process to adapt to the personalized needs of the target domain; in the joint optimization of the entire model, a curriculum learning strategy is adopted to gradually unfreeze the network layers, so that the model can adapt to the complexity of the target domain and the state migration requirements.

[0191] In some embodiments, obtaining activity data of an operator in a ship control cabin includes:

[0192] acquiring physiological data, psychological data, and behavioral data of an operator in the control cabin;

[0193] calculating a signal quality index for the physiological data, psychological data, and behavioral data, and filtering the physiological data, psychological data, and behavioral data to obtain the activity data when the signal quality index is greater than a preset threshold;

[0194] The control cabin configures the environmental adjustment parameters through an environmental adjustment actuator. After the control cabin is set using the environmental adjustment parameters, the method further includes:

[0195] Retrieving the operator's activity data;

[0196] The environmental adjustment actuator is feedback-monitored through an actuator feedback verification loop, and when it is determined that the error between the new activity data and the simulation result is greater than an error threshold, the control cabin parameters are adjusted through the environmental adjustment actuator.

[0197] Among them, dual redundant verification is established through real-time monitoring of the signal quality index (SQI) and an actuator feedback verification loop.

[0198] The SQI value is derived by comprehensively considering factors such as data validity, transmission delay, and data fluctuation to evaluate the integrity, accuracy, and consistency of physiological data, psychological data, and behavioral data.

[0199] In order to ensure that the environmental adjustment parameters match the actual execution effect, an actuator feedback verification loop is set up to perform real-time feedback monitoring of the environmental adjustment actuator in the control cabin.

[0200] By comparing the actual feedback of new activity data with the expected simulation results, the environmental control actuators are verified to be adjusting the control cabin environment according to the set parameters. If there is a significant deviation between the activity data and the simulation results, the control cabin environment is automatically corrected or an alarm is triggered, and the system enters a safe mode to ensure a stable and reliable working environment for the operator.

[0201] In the safety mode, the environmental adjustment parameters in the control cabin return to the preset safety range, the environmental adjustment operation is stopped, and guidance is provided to the operator. The ship control cabin enters a stable state that will not endanger the operator's health or the safety of the ship.

[0202] In this embodiment, by checking data in stages and providing real-time feedback monitoring of the environmental control actuator in the control cabin, the efficient, safe, and stable operation of the environmental control system in the ship's control cabin is ensured. The environmental control method for a ship's control cabin provided in this embodiment of the application can be performed by an environmental control device in the ship's control cabin. In this embodiment of the application, the environmental control device in the ship's control cabin performs the environmental control method as an example to illustrate the environmental control device in the ship's control cabin provided in this embodiment of the application.

[0203] An embodiment of the present application also provides an environmental adjustment device for a ship control cabin.

[0204] like Figure 3 As shown, the environmental adjustment device of the ship control cabin includes:

[0205] An acquisition module 310 is configured to acquire activity data of an operator in a control cabin of a ship and environmental information of the control cabin;

[0206] A first processing module 320 is configured to extract the operator's work status information from the activity data;

[0207] A second processing module 330 is configured to input the working state information and the environmental information into a parameter prediction model to obtain environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained by using a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and environmental adjustment parameter labels in a sample control cabin where the sample operator is located;

[0208] The third processing module 340 is used to call the operator's mirror model for simulation based on the environmental adjustment parameters. When the simulation result of the mirror model meets the preset simulation conditions, the control cabin is set according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics and behavioral characteristics of the operator.

[0209] According to the environmental adjustment device of the ship control cabin provided in the embodiment of the present application, by real-time monitoring of environmental information and operator activity data, the operator's fatigue and distraction can be detected in time, the environmental adjustment parameters can be verified through the operator's personalized mirror model, the operator's reaction in a specific environment is simulated, and the environmental adjustment parameters are adjusted in real time according to their physiological and psychological state. It can fully consider the environmental needs of different operators, ensure that the operator can still maintain the best working state after working for a long time, and ensure that the environmental adjustment parameters more accurately meet the needs of each operator, thereby improving the operator's work efficiency and task completion quality, as well as the overall operational efficiency of the ship control cabin, further optimizing the environment of the control cabin, not only improving the operator's working state, but also improving the safety and stability of the ship, and preventing mistakes or accidents caused by the operator's inability to adapt to the environment.

[0210] The environmental conditioning device for the ship control cabin in the embodiments of the present application may be an electronic device or a component of an electronic device, such as an integrated circuit or chip. The electronic device may be a terminal or other device other than a terminal. For example, the electronic device may be a tablet computer, a laptop computer, a PDA, a mobile internet device (MID), an ultra-mobile personal computer (UMPC), a server, or a personal computer (PC), etc., and is not specifically limited in the embodiments of the present application.

[0211] The environmental control device of the ship control cabin in the embodiment of the present application can be a device having an operating system. The operating system can be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0212] The environmental adjustment device for the ship control cabin provided in the embodiment of the present application can realize the various processes implemented in the embodiment of the environmental adjustment method for the ship control cabin in the above-mentioned embodiment. To avoid repetition, they will not be described here.

[0213] In some embodiments, as Figure 4 As shown, an embodiment of the present application also provides an electronic device 400, including a processor 401, a memory 402, and a computer program stored in the memory 402 and executable on the processor 401. When the program is executed by the processor 401, each process of the embodiment of the environmental adjustment method for the above-mentioned ship control cabin is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0214] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0215] An embodiment of the present application also provides a non-transitory computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the various processes of the above-mentioned embodiment of the environmental adjustment method for the ship control cabin and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0216] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0217] An embodiment of the present application also provides a computer program product, including a computer program, which implements the above-mentioned environmental adjustment method for the ship control cabin when executed by a processor.

[0218] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0219] An embodiment of the present application further provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned embodiment of the environmental adjustment method for the ship control cabin, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0220] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0221] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0222] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the environmental adjustment method of the ship control cabin of each embodiment of the present application.

[0223] In the description of this application, "first feature" and "second feature" may include one or more such features.

[0224] In the description of this application, “plurality” means two or more.

[0225] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

[0226] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0227] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and intent of the present application, and that the scope of the present application is defined by the claims and their equivalents.

Claims

1. A method for regulating the environment of a ship control cabin, characterized in that: include: Acquiring activity data of operators in a ship control cabin and environmental information of the control cabin; extracting the operator's work status information from the activity data; Inputting the working state information and the environmental information into a parameter prediction model to obtain the environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained through a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and an environmental adjustment parameter label in a sample control cabin where the sample operator is located; Based on the environmental adjustment parameters, the operator's mirror model is called for simulation. When the simulation result of the mirror model reaches the preset simulation conditions, the control cabin is set according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics and behavioral characteristics of the operator.

2. The environmental adjustment method for a ship control cabin according to claim 1, characterized in that: The extracting the operator's work status information from the activity data includes: aligning the activity data; Mapping the aligned activity data to a target graph structure, where nodes in the target graph structure are used to represent data points in the activity data, and edges between nodes in the target graph structure are used to represent adjacent relationships and similarities between data points; The target graph structure is feature extracted through a graph attention network to obtain the working status information.

3. The environmental adjustment method for a ship control cabin according to claim 2, characterized in that: Aligning the activity data by scale, including: Call the inference engine to reason about the active data and determine the rule category and processing agenda; Based on the rule category, calling corresponding processing rules from a rule library to match each category of data in the activity data; Invoking an execution engine to process corresponding activity data according to the processing rules based on the processing agenda, and determining sensitive data and non-sensitive data in the activity data; extracting risk assessment features from the sensitive data based on the relevance between the sensitive data and the operator's working status; The covariance matrices of the non-sensitive data and the risk assessment features are scale-aligned.

4. The environmental adjustment method for a ship control cabin according to claim 1, characterized in that: The simulation of the mirror model is based on a state transition model, which includes a domain-invariant feature extractor, a cross-modal adversarial fusion module, and an environmental parameter prediction engine connected in sequence; The step of calling the operator's mirror model to perform simulation based on the environmental adjustment parameters, and setting the control cabin according to the environmental adjustment parameters when the simulation result of the mirror model meets a preset simulation condition, includes: Performing multi-scale feature extraction on the activity data and the environmental information using the domain-invariant feature extractor to obtain a feature vector; Performing adversarial training and fusing the feature vectors through the cross-modal adversarial fusion module to generate fused features; Through the environmental parameter prediction engine, the fusion features are forward migrated to the mirror model, and the mirror model is simulated based on the environmental adjustment parameters. When it is detected that the simulation results deviate from the preset simulation conditions, reverse migration is triggered, and the parameter prediction model is optimized to obtain new environmental adjustment parameters until the simulation results meet the preset simulation conditions. The control cabin is configured according to the new environmental adjustment parameters.

5. The environmental adjustment method for a ship control cabin according to claim 4, characterized in that: The domain-invariant feature extractor includes a ResNet-18 network and a gradient reversal layer connected in sequence; The ResNet-18 network is used to perform multi-scale feature extraction on the activity data and environmental information to obtain a feature vector; The gradient reversal layer is used to perform bidirectional state migration and feature alignment on the feature vector.

6. The environmental adjustment method for a ship control cabin according to claim 4, characterized in that: When it is detected that the simulation result deviates from the preset simulation condition, triggering reverse migration and optimizing the parameter prediction model includes: When the simulation result deviates from the preset simulation condition, the incremental memory replay mechanism is triggered to calculate the historical optimal parameter combination corresponding to the activity data and the environmental information by replaying the historical data in the core memory bank and the edge memory bank; The parameter prediction model is elastically weighted and updated based on the optimal parameter combination.

7. The environmental adjustment method for a ship control cabin according to claim 4, characterized in that: Before calling the operator's mirror model to perform simulation based on the environmental adjustment parameters, the method further includes: Pre-train the state transfer model using source domain data; Freezing the domain-invariant feature extractor and the environmental parameter prediction engine, and fine-tuning the parameters of the cross-modal adversarial fusion module using target domain data; The domain-invariant feature extractor and the environmental parameter prediction engine are gradually unfrozen, and the parameters of the state transition model are optimized using target domain data.

8. The method for adjusting the environment of a ship control cabin according to any one of claims 1 to 7, characterized in that: The obtaining of activity data of operators in the ship control cabin includes: acquiring physiological data, psychological data, and behavioral data of an operator in the control cabin; calculating a signal quality index for the physiological data, psychological data, and behavioral data, and filtering the physiological data, psychological data, and behavioral data to obtain the activity data when the signal quality index is greater than a preset threshold; The control cabin configures the environmental adjustment parameters through an environmental adjustment actuator. After the control cabin is set using the environmental adjustment parameters, the method further includes: retrieving the operator's activity data; The environmental adjustment actuator is feedback-monitored through an actuator feedback verification loop, and when it is determined that the error between the new activity data and the simulation result is greater than an error threshold, the control cabin parameters are adjusted through the environmental adjustment actuator.

9. An environmental adjustment device for a ship control cabin, characterized in that: include: an acquisition module, configured to acquire activity data of an operator in a control cabin of a ship and environmental information of the control cabin; a first processing module, configured to extract the operator's working status information from the activity data; a second processing module, configured to input the working state information and the environmental information into a parameter prediction model to obtain environmental adjustment parameters output by the parameter prediction model, wherein the parameter prediction model is obtained through a plurality of training samples, the training samples including sample working state information and sample environmental information of a sample operator, and an environmental adjustment parameter label in a sample control cabin where the sample operator is located; The third processing module is used to call the operator's mirror model for simulation based on the environmental adjustment parameters. When the simulation result of the mirror model meets the preset simulation conditions, the control cabin is set according to the environmental adjustment parameters. The mirror model is constructed based on the physiological characteristics, psychological characteristics and behavioral characteristics of the operator.

10. 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 computer program, the environmental adjustment method for the ship control cabin as described in any one of claims 1-8 is implemented.

Images