Gamepad energy-saving control method and system based on intelligent analysis

Abstract

The invention discloses a gamepad energy-saving control method and system based on intelligent analysis, and the method comprises the steps: obtaining the operation monitoring information of a gamepad, and carrying out the preprocessing of the operation monitoring information of the gamepad, and obtaining the multi-mode fusion features of a target gamepad; constructing a user operation analysis model, and analyzing the operation behavior of the target user using the gamepad to obtain user operation analysis information; constructing an operation state transition matrix based on the multi-modal fusion features of the target gamepad, and predicting the operation intention of the target user; constructing an energy-saving control strategy database, acquiring energy-saving control parameters matched with the current operation scene from the energy-saving control strategy database, and establishing a dynamic regulation and control mechanism to dynamically regulate and control the energy-saving control parameters; and if the matched energy-saving control parameter combination does not exist in the energy-saving control strategy database, making an energy-saving control strategy. Intelligent energy-saving control of dynamic power consumption regulation and control is realized, so that the energy efficiency performance of the gamepad and the user experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of game controller energy-saving control, and in particular to a game controller energy-saving control method and system based on intelligent analysis. Background Art

[0002] With the rapid development of the electronic game industry, game controllers, as the core devices for human-computer interaction, have become increasingly complex in their functions, and energy consumption issues have become increasingly prominent. Traditional game controllers usually adopt a fixed power consumption mode and cannot dynamically adjust the energy consumption strategy according to the user's actual operation behavior, resulting in a large amount of energy waste. For example, when the user is browsing the menu or waiting for loading, the controller still maintains a high frequency of sensor sampling and wireless data transmission, which not only shortens the battery life, but also increases the heat of the device, affecting the user experience. In addition, the energy-saving strategies of existing controllers are mostly based on a simple timeout mechanism, lacking intelligent analysis of user operation intentions, which can easily lead to misjudgment, such as mistakenly entering the sleep state when the user pauses for a short time, resulting in delayed operation response.

[0003] At the hardware level, as the integration of game controllers increases, the number of internal sensors and sampling frequencies continue to increase, further exacerbating energy consumption pressure. For example, modern controllers are usually equipped with 9-axis IMUs, high-precision tactile sensors, and multi-channel force feedback motors. These components can consume hundreds of milliwatts of power when running at full load. However, existing energy-saving solutions mostly use simple hardware switch control, lack a fine-grained power management mechanism, and cannot dynamically adjust the working status of each module according to real-time needs. Therefore, there is an urgent need for an intelligent energy-saving control method that can deeply integrate multi-source perception data, accurately predict user intentions, and realize dynamic power consumption regulation to improve the energy efficiency performance and user experience of game controllers. Summary of the invention

[0004] The present invention overcomes the defects of the prior art and provides a game controller energy-saving control method and system based on intelligent analysis, an important purpose of which is to improve the energy efficiency performance and user experience of the game controller.

[0005] To achieve the above-mentioned purpose, the first aspect of the present invention provides a game controller energy-saving control method based on intelligent analysis, comprising: Monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller; Build a user operation analysis model, input the multimodal fusion features of the target game controller to analyze the target user's operation behavior using the game controller, and obtain user operation analysis information; The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information; Build an energy-saving control strategy database, obtain energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on user operation analysis information and operation intention prediction information, and establish a dynamic control mechanism to dynamically control the energy-saving control parameters; If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, the energy-saving control strategy is formulated based on the NSGA algorithm.

[0006] In this solution, the target game controller is monitored to obtain the controller operation monitoring information, and the controller operation monitoring information is preprocessed to obtain the multimodal fusion features of the target game controller, specifically including: A sensor array is set in the target game controller, and the target game controller is monitored by the sensor array to obtain controller operation monitoring information, wherein the controller operation monitoring information includes motion monitoring data, tactile monitoring data and environmental monitoring data; Performing time sequence processing on the handle operation monitoring information to obtain a handle operation monitoring time sequence, and using a cubic spline interpolation method to perform upsampling and alignment on the handle operation monitoring time sequence to eliminate time sequence misalignment, thereby obtaining time sequence alignment information; The timing alignment information is imported into the extended Kalman filter for dynamic correction, the motion state equation is obtained through the preset handle coordinate system, the acceleration measurement value is used as the observation, the Jacobian matrix is ​​iteratively calculated to update the attitude rotation matrix, and the corrected body acceleration is output through the updated attitude rotation matrix to eliminate the spatial deviation and obtain the time-space alignment information; Performing data cleaning processing on the spatiotemporal alignment information to eliminate outliers, and performing feature extraction based on the cleaned spatiotemporal alignment information to obtain operation monitoring features of the target game controller to form an initial feature vector; An attention mechanism is introduced to perform multimodal feature fusion, and the initial feature vector is mapped into a query vector, a key matrix and a value matrix to calculate the attention score. The calculated attention scores are weightedly fused to obtain the multimodal fusion features of the target game controller.

[0007] In this solution, the user operation analysis model is constructed, and the multimodal fusion features of the target game controller are input to analyze the operation behavior of the target user using the game controller to obtain the user operation analysis information, which specifically includes: Building a user operation analysis model based on a dual-channel neural network, the user operation analysis model includes a temporal convolution channel and a graph attention channel, wherein the temporal channel is built by a stacked temporal convolution network, and the graph attention channel is built by a graph attention network; Acquire the multimodal fusion features of the target game controller, input them into the user operation analysis model for analysis, and perform temporal feature extraction on the input multimodal fusion features through a temporal convolution channel; The input multimodal fusion features are converted into a time window sequence, and local patterns and periodic regularity are extracted from the time window sequence based on stacked multi-layer time series convolution layers to generate time series extraction features; Based on the graph attention channel, user interaction topology modeling is performed according to the input multimodal fusion features, the operation type and intensity are defined as graph nodes, and the edge weights are defined according to the time features; The cosine similarity matrix between nodes is calculated in the graph embedding space, and the soft clustering assignment matrix is ​​generated by the Sinkhorn algorithm to obtain the clustering probability. The clustering probability is spliced ​​into the node embedding as an auxiliary feature to generate an interactive topology graph. The interaction topology graph is used to obtain the adjacency matrix to characterize the target user's operation interaction characteristics, and the time series extraction features are combined for feature integration to analyze the target user's operation habits and rules to obtain user operation analysis information.

[0008] In this solution, the operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information, which specifically includes: Acquire the multimodal fusion features of the target game controller, and select the multimodal fusion features within a preset time step range based on the multimodal fusion features of the target game controller, and define them as historical multimodal fusion features; Input the historical multimodal fusion features into the user operation analysis model for analysis to obtain historical user operation analysis information, combine the historical user operation analysis information training data with the Markov algorithm to calculate the target user's operation state transition probability, and construct the operation state transition matrix; Acquire user operation analysis information, generate an operation state sequence of the target user at the current moment according to the user operation analysis information, analyze the operation state transition status of the target user at the next moment in combination with the operation state transition matrix, and generate several operation state transition sequences of the target user at the next moment; Introducing a Bayesian inference network, using the historical user operation analysis information to train the Bayesian inference network to output a transient prior distribution and a probability likelihood function, calculating the operation state sequence of the target user at the current moment, and obtaining a transient posterior distribution; The target user's operation state transition sequence at the next moment is used as the input of the trained Bayesian inference network, and the transition probability value of each operation state transition sequence is inferred in the Bayesian network according to the transient posterior distribution; By sorting the transition probability values ​​corresponding to each operation state transition sequence, the operation state transition sequence with the highest transition probability is selected as the operation intention prediction result of the target user at the next moment, and the operation intention prediction information is obtained.

[0009] In this solution, the energy-saving control strategy database is constructed, and based on the user operation analysis information and the operation intention prediction information, the energy-saving control parameters matching the current operation scenario are obtained from the energy-saving control strategy database, and a dynamic control mechanism is established to dynamically control the energy-saving control parameters, specifically including: Acquire historical game controller energy-saving control instances based on data retrieval, perform feature extraction on the historical game controller energy-saving control instances, and acquire energy-saving control parameter features of each historical game controller energy-saving control instance and corresponding historical controller operation monitoring features; Define the energy-saving control scenario type through the historical handle operation monitoring features corresponding to each historical game handle energy-saving control instance, and generate several feature subsets based on the historical handle operation monitoring features; Calculate the Euclidean distance between each feature subset and use the Euclidean distance as the subset merging index. If the Euclidean distance between two feature subsets is greater than a preset threshold, merge the corresponding two feature subsets. Obtain the merged feature subset, define the energy-saving control scenario category based on the merged feature subset, associate the energy-saving control parameter feature with the energy-saving control scenario category to obtain several association sets, and construct an energy-saving control strategy database with the energy-saving control scenario type-energy-saving control scenario type feature-energy-saving control parameter as the path based on the several association sets; Obtaining user operation analysis information and operation intention prediction information, and generating a handle usage portrait of the target user at the current moment according to the user operation analysis information and operation intention prediction information; Import the target game controller usage profile at the current moment into the energy-saving control strategy database, use the cosine metric algorithm to match the energy-saving control scene type, and select the energy-saving control parameters of the energy-saving control scene type with the highest similarity as the final energy-saving management and control plan, and generate control instructions to perform energy-saving control on the target game controller; A dynamic control mechanism is established to analyze the difference between the operation intention prediction results and the real-time operation characteristics under real-time conditions by monitoring the operation data of the target handle in real time. If the difference is greater than the preset difference threshold, the handle usage portrait of the target user is corrected according to the real-time operation characteristics, and new energy-saving control parameters are dynamically selected based on the corrected handle usage portrait of the target user.

[0010] In this solution, if there is no matching energy-saving control parameter combination in the energy-saving control strategy database, an energy-saving control strategy is formulated based on the NSGA algorithm, specifically including: If the energy-saving control strategy database fails to match the energy-saving control scenario that matches the handle usage profile of the target user at the current moment, then the three energy-saving control parameter combinations with the highest similarity between the current energy-saving control strategy database and the handle usage profile of the target user at the current moment are obtained to formulate an energy-saving control strategy; Based on the obtained three energy-saving control parameter combinations, an energy-saving control parameter interval is generated, a search space is formed according to the energy-saving control parameter interval, and the NSGA algorithm is introduced to generate an initial population based on the search space; Preset the objective function and constraint conditions, calculate the fitness value of each individual in the initial population through the objective function, sort all individuals in the population according to the fitness value, and obtain several non-dominated layers; The crowding distance is calculated for the individuals in each non-dominated layer, and the elite individuals are selected by the tournament selection method using the non-dominated sorting level and crowding distance, and the selected elite individuals are subjected to crossover and mutation operations. Merge the parent generation with the child generation, recalculate the non-dominated sorting and crowding distance of all candidate solutions, and select a preset number of elite individuals according to the Pareto frontier level and crowding distance to form the next generation population for iterative evolution; When the convergence criteria are met or the preset number of iterations is reached, a non-dominated solution set is output, and the energy-saving control parameter combination at the current moment is generated according to the non-dominated solution set, and a control instruction is generated to control the target game controller.

[0011] A second aspect of the present invention provides a game controller energy-saving control system based on intelligent analysis, the system comprising: a memory, a processor, the memory containing a game controller energy-saving control method program based on intelligent analysis, the game controller energy-saving control method program based on intelligent analysis when executed by the processor to implement the following steps: Monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller; Build a user operation analysis model, input the multimodal fusion features of the target game controller to analyze the target user's operation behavior using the game controller, and obtain user operation analysis information; The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information; Build an energy-saving control strategy database, obtain energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on user operation analysis information and operation intention prediction information, and establish a dynamic control mechanism to dynamically control the energy-saving control parameters; If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, the energy-saving control strategy is formulated based on the NSGA algorithm.

[0012] The present invention discloses a game controller energy-saving control method and system based on intelligent analysis, including: obtaining the operation monitoring information of the controller, pre-processing the operation monitoring information of the controller to obtain the multi-modal fusion characteristics of the target game controller; constructing a user operation analysis model, analyzing the operation behavior of the target user using the game controller, and obtaining the user operation analysis information; constructing an operation state transfer matrix based on the multi-modal fusion characteristics of the target game controller, and predicting the operation intention of the target user; constructing an energy-saving control strategy database, obtaining energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database, and establishing a dynamic control mechanism to dynamically control the energy-saving control parameters; if there is no matching energy-saving control parameter combination in the energy-saving control strategy database, then formulating an energy-saving control strategy. Intelligent energy-saving control with dynamic power consumption regulation is realized to improve the energy efficiency performance and user experience of the game controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments or exemplary embodiments of the present invention, the drawings required for use in the embodiments or exemplary descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained according to the drawings without paying creative work.

[0014] Figure 1 A flow chart of a game controller energy saving control method based on intelligent analysis provided by one embodiment of the present invention; Figure 2 A flow chart of a method for formulating a game controller energy-saving control strategy provided by an embodiment of the present invention; Figure 3 A block diagram of a game controller energy-saving control system based on intelligent analysis provided by one embodiment of the present invention; The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0015] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0016] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0017] Figure 1 A flow chart of a game controller energy saving control method based on intelligent analysis provided by one embodiment of the present invention; like Figure 1 As shown, the present invention provides a flow chart of a game controller energy-saving control method based on intelligent analysis, comprising: S102, monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain a multimodal fusion feature of the target game controller; S104, constructing a user operation analysis model, inputting the multimodal fusion features of the target game controller to analyze the operation behavior of the target user using the game controller, and obtaining user operation analysis information; S106, constructing an operation state transfer matrix based on the multimodal fusion features of the target game controller, and predicting the operation intention of the target user in combination with a Bayesian inference network to obtain operation intention prediction information; S108, building an energy-saving control strategy database, acquiring energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on the user operation analysis information and the operation intention prediction information, and establishing a dynamic control mechanism to dynamically control the energy-saving control parameters; S110: If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, an energy-saving control strategy is formulated based on the NSGA algorithm.

[0018] Further, in a preferred embodiment of the present invention, the monitoring of the target game controller to obtain the controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain the multimodal fusion features of the target game controller specifically include: A sensor array is set in the target game controller, and the target game controller is monitored by the sensor array to obtain controller operation monitoring information, wherein the controller operation monitoring information includes motion monitoring data, tactile monitoring data and environmental monitoring data; Performing time sequence processing on the handle operation monitoring information to obtain a handle operation monitoring time sequence, and using a cubic spline interpolation method to perform upsampling and alignment on the handle operation monitoring time sequence to eliminate time sequence misalignment, thereby obtaining time sequence alignment information; The timing alignment information is imported into the extended Kalman filter for dynamic correction, the motion state equation is obtained through the preset handle coordinate system, the acceleration measurement value is used as the observation, the Jacobian matrix is ​​iteratively calculated to update the attitude rotation matrix, and the corrected body acceleration is output through the updated attitude rotation matrix to eliminate the spatial deviation and obtain the time-space alignment information; Performing data cleaning processing on the spatiotemporal alignment information to eliminate outliers, and performing feature extraction based on the cleaned spatiotemporal alignment information to obtain operation monitoring features of the target game controller to form an initial feature vector; An attention mechanism is introduced to perform multimodal feature fusion, and the initial feature vector is mapped into a query vector, a key matrix and a value matrix to calculate the attention score. The calculated attention scores are weightedly fused to obtain the multimodal fusion features of the target game controller.

[0019] It should be noted that deploying a sensor array in the target game controller is the basic architecture for achieving refined energy consumption control. The array usually includes a six-axis inertial measurement unit (accelerometer + gyroscope), a linear resonant actuator (LRA) tactile sensor, a surface capacitive pressure sensor and other multi-sensor devices. Motion monitoring data covers physical quantities such as three-dimensional acceleration vector, angular velocity change rate and attitude Euler angle. Tactile monitoring data includes interactive parameters such as key trigger force gradient, vibration motor current waveform and tactile feedback frequency. Environmental monitoring data involves peripheral information such as handle surface temperature distribution, grip area capacitance change and ambient light intensity. These heterogeneous data streams are generated asynchronously at different sampling frequencies (such as 100Hz for accelerometer and 1kHz for tactile feedback), and hardware interrupt coordination and DMA transfer mechanisms are required to ensure the complete capture of the original monitoring data.

[0020] It should be noted that the timing alignment process aims to eliminate the sampling time difference and clock drift problems of multi-source sensors. After mapping the raw data streams of each sensor to a unified timing coordinate system according to the physical timestamp, the cubic spline interpolation method is used for upsampling processing for low-frequency data channels (such as 10Hz for temperature sensors). Its mathematical essence is to construct a piecewise cubic polynomial function to meet the continuity constraints of the function value, first-order derivative and second-order derivative at adjacent data points, and increase the sampling rate to synchronize with the highest frequency sensor (such as 1kHz) while maintaining data smoothness. This process minimizes the integral square error between the interpolation curve and the original data to ensure that transient features such as high-frequency vibration signals are not over-smoothed. Dynamic correction based on the extended Kalman filter (EKF) is the core link to solve nonlinear errors in sensor fusion. The quaternion attitude motion model is established based on the handle body coordinate system. The state vector includes position, velocity, attitude quaternion and gyroscope bias. By calculating the Jacobian determinant of the state transfer matrix in real time, the nonlinear motion equation is linearized at the latest estimated point, and the covariance matrix is ​​iteratively updated. In particular, a compensation mechanism based on gravity vector observation is designed to address the centrifugal acceleration interference in accelerometer data: when the handle is in a non-free motion state (such as a sudden swing by the user), it automatically switches to the gyroscope-dominated angular velocity integration mode to avoid the accumulation of inertial navigation errors. In the data cleaning stage, the improved Grubbs test and DBSCAN clustering algorithm can be used to identify global outliers (such as acceleration spikes caused by transient sensor failures), and then density clustering can be used to eliminate outliers with incoherent local contexts (such as abnormal vibration signals that appear in a static state). The multimodal attention fusion mechanism maps the initial feature vector to a 128-dimensional query vector space through a learnable feature projection matrix, and constructs a cross-modal key-value matrix to capture the implicit associations between sensors. A gated cross-attention layer is designed to consider the dot product similarity between features when calculating the significance weight of tactile features to motion features. The resulting fused feature vector provides a highly discriminative feature expression for dynamically adjusting the sensor power supply strategy.

[0021] Further, in a preferred embodiment of the present invention, the user operation analysis model is constructed, and the multimodal fusion features of the target game controller are input to analyze the operation behavior of the target user using the game controller to obtain the user operation analysis information, which specifically includes: Building a user operation analysis model based on a dual-channel neural network, the user operation analysis model includes a temporal convolution channel and a graph attention channel, wherein the temporal channel is built by a stacked temporal convolution network, and the graph attention channel is built by a graph attention network; Acquire the multimodal fusion features of the target game controller, input them into the user operation analysis model for analysis, and perform temporal feature extraction on the input multimodal fusion features through a temporal convolution channel; The input multimodal fusion features are converted into a time window sequence, and local patterns and periodic regularity are extracted from the time window sequence based on stacked multi-layer time series convolution layers to generate time series extraction features; Based on the graph attention channel, user interaction topology modeling is performed according to the input multimodal fusion features, the operation type and intensity are defined as graph nodes, and the edge weights are defined according to the time features; The cosine similarity matrix between nodes is calculated in the graph embedding space, and the soft clustering assignment matrix is ​​generated by the Sinkhorn algorithm to obtain the clustering probability. The clustering probability is spliced ​​into the node embedding as an auxiliary feature to generate an interactive topology graph. The interaction topology graph is used to obtain the adjacency matrix to characterize the target user's operation interaction characteristics, and the time series extraction features are combined for feature integration to analyze the target user's operation habits and rules to obtain user operation analysis information.

[0022] It should be noted that the user operation analysis model based on the dual-channel neural network is intended to collaboratively mine the temporal regularity and interactive topological characteristics of user operation behaviors. The temporal convolution channel uses a stacked dilated causal convolutional network as the core structure. Its multi-layer dilated convolution kernel gradually expands the temporal receptive field through exponential expansion factors (such as 1, 2, 4, 8). The bottom convolution layer captures the micro-operation rhythm, and the high-level network identifies the macro-tactical cycle. The input multimodal fusion features are first divided into a fixed-length time window sequence (such as a 256-frame window with a frame interval of 10ms). After the data in each window is zero-phase filtered and standardized, the propagation intensity of the feature flow at different time scales is controlled by the gated linear unit (GLU). Finally, the temporal feature vector containing the local operation mode signature (such as the joystick steering acceleration waveform) and the global behavior cycle identifier (such as the shooting frequency Fourier coefficient) is output. The parallel graph attention channel focuses on modeling the complex association network between operation events. The discrete operation events (such as button triggering and joystick deflection) within each time window are abstracted into graph nodes. The node attributes include the operation type encoding and the strength quantization value (normalized pressure, displacement, etc.). The edge weights between nodes are dynamically determined by two factors: the exponential decay function based on the operation interval in the time dimension reflects the temporal closeness between operations, and the pre-trained operation transfer probability in the semantic dimension describes the tactical logic association. The graph attention network adopts a multi-head mechanism. Each attention head learns a different interaction mode projection space. By calculating the adaptive attention coefficient between nodes, the neighborhood features are aggregated to generate node embeddings rich in interaction semantics.

[0023] It should be noted that in order to further reveal the intrinsic pattern structure of the operation behavior, a node similarity matrix is ​​constructed in the graph embedding space, and a differentiable clustering technology based on the Sinkhorn-Knopp algorithm is used. The algorithm inputs the cosine similarity matrix between nodes into the double random matrix constraint space, and generates a soft clustering assignment matrix through iterative row and column normalization operations, so that each node obtains a probability distribution on different tactical mode categories. The clustering probability is used as an auxiliary feature, and after being spliced ​​with the original node embedding vector, the position encoding information is injected through a nonlinear transformation to form an enhanced graph representation. In the final feature integration stage, a cross-modal gated fusion mechanism is designed to adaptively weight the rhythm features extracted by the temporal convolution channel and the topological features generated by the graph attention channel. Based on the fused feature vector, the core rules of user operations are parsed, including the operation intensity distribution histogram, key combination rules, etc., to provide a quantitative basis for personalized energy efficiency regulation.

[0024] Furthermore, in a preferred embodiment of the present invention, the operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information, which specifically includes: Acquire the multimodal fusion features of the target game controller, and select the multimodal fusion features within a preset time step range based on the multimodal fusion features of the target game controller, and define them as historical multimodal fusion features; Input the historical multimodal fusion features into the user operation analysis model for analysis to obtain historical user operation analysis information, combine the historical user operation analysis information training data with the Markov algorithm to calculate the target user's operation state transition probability, and construct the operation state transition matrix; Acquire user operation analysis information, generate an operation state sequence of the target user at the current moment according to the user operation analysis information, analyze the operation state transition status of the target user at the next moment in combination with the operation state transition matrix, and generate several operation state transition sequences of the target user at the next moment; Introducing a Bayesian inference network, using the historical user operation analysis information to train the Bayesian inference network to output a transient prior distribution and a probability likelihood function, calculating the operation state sequence of the target user at the current moment, and obtaining a transient posterior distribution; The target user's operation state transition sequence at the next moment is used as the input of the trained Bayesian inference network, and the transition probability value of each operation state transition sequence is inferred in the Bayesian network according to the transient posterior distribution; By sorting the transition probability values ​​corresponding to each operation state transition sequence, the operation state transition sequence with the highest transition probability is selected as the operation intention prediction result of the target user at the next moment, and the operation intention prediction information is obtained.

[0025] It should be noted that the operation intention prediction achieves accurate prediction by integrating historical behavior analysis and real-time state reasoning. First, the historical data within the preset time window is intercepted from the multimodal fusion features, that is, the data before the preset time window. After normalization, it forms a tensor structure containing multidimensional parameters such as motion trajectory, key pressure and device power consumption, which is input into the user operation analysis model for deep feature analysis. The temporal convolution module extracts the rhythm pattern and periodic law in the user operation, and the graph attention network constructs the correlation topology between operation events (such as the strong causal relationship between "rolling" and "shooting"), and outputs the feature vector representing the user's key habits and operation intensity preference as the basis for Markov state modeling. Based on the feature vector cluster obtained by analysis, the sliding window clustering algorithm is used to divide the discrete operation state space, and each state node encodes the multidimensional feature distribution of a specific operation mode (such as "precise sniping" and "high-speed movement"). The Markov transition probability matrix is ​​constructed by statistically analyzing the historical state transition frequency, in which the transition weight between states not only considers the explicit operation sequence, but also introduces the temporal attenuation factor (such as giving higher weight to the transfer records within the last 10 minutes) to enhance the predictive influence of recent behavior patterns. After obtaining the current operation state sequence, a multi-step prediction path search is carried out in the transfer matrix, and the Viterbi algorithm is used to screen out N (N=5) maximum likelihood paths. Each path corresponds to a possible operation evolution trajectory in the future Δt time, and its basic confidence is determined by the continuous product of the state transition probabilities on the path. In order to enhance the contextual adaptability of the prediction, a Bayesian network is constructed to obtain the state transition likelihood function through variational inference training, and the conditional probability table is dynamically updated based on real-time sensor data. The Markov path set is input into the Bayesian network for probability correction, and the posterior probability of each path in the current context is calculated to ensure that short-term predictions are more dependent on operation inertia, and long-term predictions tend to general modes, and the path with the highest comprehensive confidence is selected as the intention prediction result.

[0026] Further, in a preferred embodiment of the present invention, the energy-saving control strategy database is constructed, and based on the user operation analysis information and the operation intention prediction information, the energy-saving control parameters matching the current operation scenario are obtained in the energy-saving control strategy database, and a dynamic control mechanism is established to dynamically control the energy-saving control parameters, specifically including: Acquire historical game controller energy-saving control instances based on data retrieval, perform feature extraction on the historical game controller energy-saving control instances, and acquire energy-saving control parameter features of each historical game controller energy-saving control instance and corresponding historical controller operation monitoring features; Define the energy-saving control scenario type through the historical handle operation monitoring features corresponding to each historical game handle energy-saving control instance, and generate several feature subsets based on the historical handle operation monitoring features; Calculate the Euclidean distance between each feature subset and use the Euclidean distance as the subset merging index. If the Euclidean distance between two feature subsets is greater than a preset threshold, merge the corresponding two feature subsets. Obtain the merged feature subset, define the energy-saving control scenario category based on the merged feature subset, associate the energy-saving control parameter feature with the energy-saving control scenario category to obtain several association sets, and construct an energy-saving control strategy database with the energy-saving control scenario type-energy-saving control scenario type feature-energy-saving control parameter as the path based on the several association sets; Obtaining user operation analysis information and operation intention prediction information, and generating a handle usage portrait of the target user at the current moment according to the user operation analysis information and operation intention prediction information; Import the target game controller usage profile at the current moment into the energy-saving control strategy database, use the cosine metric algorithm to match the energy-saving control scene type, and select the energy-saving control parameters of the energy-saving control scene type with the highest similarity as the final energy-saving management and control plan, and generate control instructions to perform energy-saving control on the target game controller; A dynamic control mechanism is established to analyze the difference between the operation intention prediction results and the real-time operation characteristics under real-time conditions by monitoring the operation data of the target handle in real time. If the difference is greater than the preset difference threshold, the handle usage portrait of the target user is corrected according to the real-time operation characteristics, and new energy-saving control parameters are dynamically selected based on the corrected handle usage portrait of the target user.

[0027] It should be noted that adaptive control is achieved by building a historical strategy database and a real-time portrait matching mechanism. First, multi-dimensional energy-saving control instance data is extracted from the historical operation log. Each instance contains a set of control parameters (such as gyroscope sampling rate, tactile feedback intensity level) and a corresponding operation monitoring feature matrix. The latter is composed of heterogeneous data such as six-axis inertial sensor waveforms, key pressure gradient spectra, and power consumption timing curves, forming a normalized feature vector to represent the unique pattern of each historical control case. The scene classification based on feature similarity adopts a hierarchical clustering algorithm. Initially, each historical instance is regarded as an independent feature subset, and the Euclidean distance between subsets is calculated. When the subset spacing exceeds the preset threshold, the merge operation is triggered. The merge process retains the feature distribution statistics of each subset, and the merged feature distribution is refitted through a Gaussian mixture model. In the final set of scene categories, each category corresponds to a specific operation mode and its typical energy consumption feature spectrum, and a strong correlation mapping is established with the historical successful control parameters (such as reducing the power consumption by 38% when the gyroscope is reduced from 100Hz to 60Hz). Based on the user operation analysis information and operation intention prediction information, the target user's handle usage portrait is constructed at the current moment. The multi-dimensional features (operation frequency 32Hz, steering smoothness 0.87, etc.) are encoded into a 64-dimensional latent space vector through a deep feature projection network. The strategy matching engine uses a cosine similarity calculation framework. When searching for scene categories in the latent space, a 3x weight is applied to feature dimensions with significant control effects (such as the slope of the power consumption curve) to amplify their influence.

[0028] It should be noted that the dynamic control mechanism detects the morphological difference between the actual operation flow and the predicted sequence through the sliding window DTW distance (window length 128 frames) in the short term, and evaluates the distribution deviation of the user portrait through the KL divergence in the long term. When the comprehensive difference index exceeds the threshold, the incremental learning process is triggered: first freeze the current policy executor, then fine-tune the feature projection network parameters with the latest 500 frames of data, update the weights of the fully connected layer, and then map the new operation mode to the existing scene category or create a new category through the fast clustering algorithm, and then match the corresponding energy-saving control scheme.

[0029] Figure 2 A flow chart of a method for formulating a game controller energy-saving control strategy provided by an embodiment of the present invention; like Figure 2 As shown, the present invention provides a flow chart of a method for formulating a game controller energy-saving control strategy, comprising: S202, if the energy-saving control strategy database fails to match the energy-saving control scenario that matches the handle usage portrait of the target user at the current moment, then obtain the three energy-saving control parameter combinations with the highest similarity between the current energy-saving control strategy database and the handle usage portrait of the target user at the current moment, and formulate an energy-saving control strategy; S204, generating an energy-saving control parameter interval based on the obtained three energy-saving control parameter combinations, forming a search space according to the energy-saving control parameter interval, and introducing an NSGA algorithm to generate an initial population based on the search space; S206, preset an objective function and constraint conditions, calculate the fitness value of each individual in the initial population through the objective function, perform non-dominated sorting on all individuals in the population according to the fitness value, and obtain a number of non-dominated layers; S208, calculating the crowding distance for the individuals in each non-dominated layer, selecting elite individuals using the tournament selection method using the non-dominated sorting level and the crowding distance, and performing crossover and mutation operations on the selected elite individuals; S210, merging the parent generation and the child generation, recalculating the non-dominated sorting and crowding distance of all candidate solutions, and selecting a preset number of elite individuals according to the Pareto frontier level and crowding distance to form the next generation population for iterative evolution; S212, when the convergence standard is met or the preset number of iterations is reached, a non-dominated solution set is output, and the energy-saving control parameter combination at the current moment is generated according to the non-dominated solution set, and a control instruction is generated to control the target game controller.

[0030] It should be noted that in the scenario where the historical policy library and the current user portrait do not match, the multi-objective evolutionary optimization process is started to generate an adaptability control strategy. First, the three historical parameter combinations with the highest cosine similarity to the current handle usage portrait are recalled from the database, and a multi-dimensional continuous search space is constructed by extracting the upper and lower boundaries of each parameter. For example, the gyroscope frequency range [40,60] Hz forms a continuous solution domain, and the discrete level of tactile intensity is mapped to a linear interpolation interval of [20%, 45%]. The initial population is generated based on sampling in the search space to ensure that individuals evenly cover the possible solution area. A dual-objective optimization function is defined: the first objective is the energy efficiency improvement rate, which is calculated as the ratio of the integral of the power consumption curve under the current strategy to the baseline strategy; the second objective is the operation quality loss, which is quantified by the dynamic time warping (DTW) distance between the predicted intention sequence and the actual response. Constraints include hardware safety thresholds (such as temperature not exceeding 45°C) and user experience red lines (such as operation delay <25ms). The dual-objective fitness value of each individual is obtained, and the non-dominated sorting algorithm is used to divide the population into multiple frontier levels. The first front contains all Pareto optimal solutions that are not dominated by other solutions, the second front is the solution set dominated only by the first front, and so on.

[0031] It should be noted that in the elite selection stage, the crowding distance index is calculated for individuals in each non-dominated level. This index ensures the diversity of the frontier by measuring the density of the solution with the adjacent solutions in the target space. A binary tournament selection mechanism is adopted: two individuals are randomly selected, and the one with a higher frontier level is preferred. If the levels are the same, the one with a larger crowding distance is selected. The selected elite individuals are simulated binary crossover and polynomial mutation, where the crossover probability is set to 0.9, the mutation probability is 0.1, and the distribution index is adjusted to 5 to balance exploration and development. During the iterative optimization process, the parent population and the offspring are merged to form a temporary population (size 400), and the non-dominated sorting and crowding distance calculation are re-executed, and the top 200 optimal individuals are retained to form a new generation of population. Set the dynamic convergence judgment standard: the evolution is terminated when the change rate of the hypervolume index of the frontier solution set for 5 consecutive generations is less than 1% or reaches the upper limit of 50 generations of iterations. The non-dominated solution set finally output selects the optimal control parameter combination for regulation.

[0032] Furthermore, the objective function can be: ; in, is the weight coefficient used to balance the optimization priority between energy efficiency and operation quality. is the energy efficiency improvement rate, that is, the power consumption optimization effect, is the operation quality loss degree, which is constructed based on the average operation response delay of the current strategy and the benchmark strategy and the false trigger rate of the current strategy and the benchmark strategy.

[0033] Figure 3 A game controller energy-saving control system 3 based on intelligent analysis is provided in one embodiment of the present invention. The system comprises: a memory 31 and a processor 32. The memory 31 contains a game controller energy-saving control method program based on intelligent analysis. When the game controller energy-saving control method program based on intelligent analysis is executed by the processor 32, the following steps are implemented: Monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller; Build a user operation analysis model, input the multimodal fusion features of the target game controller to analyze the target user's operation behavior using the game controller, and obtain user operation analysis information; The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information; Build an energy-saving control strategy database, obtain energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on user operation analysis information and operation intention prediction information, and establish a dynamic control mechanism to dynamically control the energy-saving control parameters; If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, the energy-saving control strategy is formulated based on the NSGA algorithm.

[0034] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0035] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0036] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0037] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0038] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0039] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A game controller energy-saving control method based on intelligent analysis, characterized in that: include: Monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller; Build a user operation analysis model, input the multimodal fusion features of the target game controller to analyze the target user's operation behavior using the game controller, and obtain user operation analysis information; The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information; Build an energy-saving control strategy database, obtain energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on user operation analysis information and operation intention prediction information, and establish a dynamic control mechanism to dynamically control the energy-saving control parameters; If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, the energy-saving control strategy is formulated based on the NSGA algorithm.

2. The game controller energy saving control method based on intelligent analysis according to claim 1 is characterized in that: The step of monitoring the target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller specifically includes: A sensor array is set in the target game controller, and the target game controller is monitored by the sensor array to obtain controller operation monitoring information, wherein the controller operation monitoring information includes motion monitoring data, tactile monitoring data and environmental monitoring data; Performing time sequence processing on the handle operation monitoring information to obtain a handle operation monitoring time sequence, and using a cubic spline interpolation method to perform upsampling and alignment on the handle operation monitoring time sequence to eliminate time sequence misalignment, thereby obtaining time sequence alignment information; The timing alignment information is imported into the extended Kalman filter for dynamic correction, the motion state equation is obtained through the preset handle coordinate system, the acceleration measurement value is used as the observation, the Jacobian matrix is ​​iteratively calculated to update the attitude rotation matrix, and the corrected body acceleration is output through the updated attitude rotation matrix to eliminate the spatial deviation and obtain the time-space alignment information; Performing data cleaning processing on the spatiotemporal alignment information to eliminate outliers, and performing feature extraction based on the cleaned spatiotemporal alignment information to obtain operation monitoring features of the target game controller to form an initial feature vector; An attention mechanism is introduced to perform multimodal feature fusion, and the initial feature vector is mapped into a query vector, a key matrix and a value matrix to calculate the attention score. The calculated attention scores are weightedly fused to obtain the multimodal fusion features of the target game controller.

3. The game controller energy saving control method based on intelligent analysis according to claim 1, characterized in that: The user operation analysis model is constructed, and the multimodal fusion features of the target game controller are input to analyze the operation behavior of the target user using the game controller to obtain the user operation analysis information, which specifically includes: Building a user operation analysis model based on a dual-channel neural network, the user operation analysis model includes a temporal convolution channel and a graph attention channel, wherein the temporal channel is built by a stacked temporal convolution network, and the graph attention channel is built by a graph attention network; Acquire the multimodal fusion features of the target game controller, input them into the user operation analysis model for analysis, and perform temporal feature extraction on the input multimodal fusion features through a temporal convolution channel; The input multimodal fusion features are converted into a time window sequence, and local patterns and periodic regularity are extracted from the time window sequence based on stacked multi-layer time series convolution layers to generate time series extraction features; Based on the graph attention channel, user interaction topology modeling is performed according to the input multimodal fusion features, the operation type and intensity are defined as graph nodes, and the edge weights are defined according to the time features; The cosine similarity matrix between nodes is calculated in the graph embedding space, and the soft clustering assignment matrix is ​​generated by the Sinkhorn algorithm to obtain the clustering probability. The clustering probability is spliced ​​into the node embedding as an auxiliary feature to generate an interactive topology graph. The interaction topology graph is used to obtain the adjacency matrix to characterize the target user's operation interaction characteristics, and the time series extraction features are combined for feature integration to analyze the target user's operation habits and rules to obtain user operation analysis information.

4. The game controller energy saving control method based on intelligent analysis according to claim 1, characterized in that: The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information, which specifically includes: Acquire the multimodal fusion features of the target game controller, and select the multimodal fusion features within a preset time step range based on the multimodal fusion features of the target game controller, and define them as historical multimodal fusion features; Input the historical multimodal fusion features into the user operation analysis model for analysis to obtain historical user operation analysis information, combine the historical user operation analysis information training data with the Markov algorithm to calculate the target user's operation state transition probability, and construct the operation state transition matrix; Acquire user operation analysis information, generate an operation state sequence of the target user at the current moment according to the user operation analysis information, analyze the operation state transition status of the target user at the next moment in combination with the operation state transition matrix, and generate several operation state transition sequences of the target user at the next moment; Introducing a Bayesian inference network, using the historical user operation analysis information to train the Bayesian inference network to output a transient prior distribution and a probability likelihood function, calculating the operation state sequence of the target user at the current moment, and obtaining a transient posterior distribution; The target user's operation state transition sequence at the next moment is used as the input of the trained Bayesian inference network, and the transition probability value of each operation state transition sequence is inferred in the Bayesian network according to the transient posterior distribution; By sorting the transition probability values ​​corresponding to each operation state transition sequence, the operation state transition sequence with the highest transition probability is selected as the operation intention prediction result of the target user at the next moment, and the operation intention prediction information is obtained.

5. The game controller energy saving control method based on intelligent analysis according to claim 1 is characterized in that: The energy-saving control strategy database is constructed, and based on the user operation analysis information and the operation intention prediction information, energy-saving control parameters matching the current operation scenario are obtained from the energy-saving control strategy database, and a dynamic control mechanism is established to dynamically control the energy-saving control parameters, specifically including: Acquire historical game controller energy-saving control instances based on data retrieval, perform feature extraction on the historical game controller energy-saving control instances, and acquire energy-saving control parameter features of each historical game controller energy-saving control instance and corresponding historical controller operation monitoring features; Define the energy-saving control scenario type through the historical handle operation monitoring features corresponding to each historical game handle energy-saving control instance, and generate several feature subsets based on the historical handle operation monitoring features; Calculate the Euclidean distance between each feature subset and use the Euclidean distance as the subset merging index. If the Euclidean distance between two feature subsets is greater than a preset threshold, merge the corresponding two feature subsets. Obtain the merged feature subset, define the energy-saving control scenario category based on the merged feature subset, associate the energy-saving control parameter feature with the energy-saving control scenario category to obtain several association sets, and construct an energy-saving control strategy database with the energy-saving control scenario type-energy-saving control scenario type feature-energy-saving control parameter as the path based on the several association sets; Obtaining user operation analysis information and operation intention prediction information, and generating a handle usage portrait of the target user at the current moment according to the user operation analysis information and operation intention prediction information; Import the target game controller usage profile at the current moment into the energy-saving control strategy database, use the cosine metric algorithm to match the energy-saving control scene type, and select the energy-saving control parameters of the energy-saving control scene type with the highest similarity as the final energy-saving management and control plan, and generate control instructions to perform energy-saving control on the target game controller; A dynamic control mechanism is established to analyze the difference between the operation intention prediction results and the real-time operation characteristics under real-time conditions by monitoring the operation data of the target handle in real time. If the difference is greater than the preset difference threshold, the handle usage portrait of the target user is corrected according to the real-time operation characteristics, and new energy-saving control parameters are dynamically selected based on the corrected handle usage portrait of the target user.

6. The game controller energy saving control method based on intelligent analysis according to claim 1, characterized in that: If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, an energy-saving control strategy is formulated based on the NSGA algorithm, specifically including: If the energy-saving control strategy database fails to match the energy-saving control scenario that matches the handle usage profile of the target user at the current moment, then the three energy-saving control parameter combinations with the highest similarity between the current energy-saving control strategy database and the handle usage profile of the target user at the current moment are obtained to formulate an energy-saving control strategy; Based on the obtained three energy-saving control parameter combinations, an energy-saving control parameter interval is generated, a search space is formed according to the energy-saving control parameter interval, and the NSGA algorithm is introduced to generate an initial population based on the search space; Preset the objective function and constraint conditions, calculate the fitness value of each individual in the initial population through the objective function, sort all individuals in the population according to the fitness value, and obtain several non-dominated layers; The crowding distance is calculated for the individuals in each non-dominated layer, and the elite individuals are selected by the tournament selection method using the non-dominated sorting level and crowding distance, and the selected elite individuals are subjected to crossover and mutation operations. Merge the parent generation with the child generation, recalculate the non-dominated sorting and crowding distance of all candidate solutions, and select a preset number of elite individuals according to the Pareto frontier level and crowding distance to form the next generation population for iterative evolution; When the convergence criteria are met or the preset number of iterations is reached, a non-dominated solution set is output, and the energy-saving control parameter combination at the current moment is generated according to the non-dominated solution set, and a control instruction is generated to control the target game controller.

7. A game controller energy-saving control system based on intelligent analysis, characterized in that: The system includes: a memory and a processor, wherein the memory contains a game controller energy-saving control method program based on intelligent analysis, and when the game controller energy-saving control method program based on intelligent analysis is executed by the processor, the following steps are implemented: Monitoring a target game controller to obtain controller operation monitoring information, and preprocessing the controller operation monitoring information to obtain multimodal fusion features of the target game controller; Build a user operation analysis model, input the multimodal fusion features of the target game controller to analyze the target user's operation behavior using the game controller, and obtain user operation analysis information; The operation state transfer matrix is ​​constructed based on the multimodal fusion features of the target game controller, and the operation intention of the target user is predicted by combining the Bayesian inference network to obtain the operation intention prediction information; Build an energy-saving control strategy database, obtain energy-saving control parameters matching the current operation scenario in the energy-saving control strategy database based on user operation analysis information and operation intention prediction information, and establish a dynamic control mechanism to dynamically control the energy-saving control parameters; If there is no matching energy-saving control parameter combination in the energy-saving control strategy database, the energy-saving control strategy is formulated based on the NSGA algorithm.

Images