Intelligent game interaction method for intelligent park

Abstract

The invention relates to the technical field of intelligent interaction, and discloses an intelligent game interaction method for an intelligent park, which comprises the following steps of: 1, acquiring behavior data of players through a plurality of sensors; step 2, constructing a dynamic Hamiltonian optimization model containing a physiological signal coupling item, and solving the quantum state representation through a quantum optimization algorithm to generate a multi-target optimization parameter; 3, based on the optimization parameters, carrying out cooperative adjustment on music rhythm and task difficulty in game interaction; and step 4, performing incremental updating on the dynamic Hamiltonian optimization model according to the adjusted player feedback data. According to the method, the space-time registration algorithm of the millimeter wave radar, the visual sensor and the biological signal is adopted, high-fidelity representation of the human body motion mode in the quantum Hilbert space is achieved, and quantization input with definite physical significance is provided for subsequent optimization.

Description

Technical Field

[0001] The present invention relates to the field of intelligent interaction technology, and in particular to an intelligent game interaction method for a smart park. Background Art

[0002] With the advancement of smart city construction, smart parks, as an important carrier of the intelligent transformation of urban public spaces, have upgraded their core functions from basic services to immersive interactive experiences. The intelligent game interactive system aims to provide tourists with digital entertainment services that are both interesting and healthy and educational by integrating the Internet of Things, augmented reality and somatosensory interaction technologies. Existing technologies usually use multimodal sensors to collect user behavior data, combined with a preset rule engine to adjust game parameters, and try to maintain a balance between player participation and challenge in dynamic interaction.

[0003] The current mainstream intelligent game systems mostly rely on motion capture solutions using visual sensors and inertial measurement units (IMUs), supplemented by monitoring of basic physiological indicators such as heart rate and electromyography. Although such systems can achieve basic motion trajectory tracking and simple physiological feedback, they have problems such as limited accuracy in motion feature extraction and poor spatiotemporal consistency of multi-source data in complex scenarios. Especially in multi-player collaborative interaction scenarios, traditional methods are difficult to effectively distinguish individual action intentions, resulting in delayed feedback and distorted experience. In addition, existing parameter adjustment mechanisms mostly use threshold-triggered rules, lack the ability to continuously model the dynamic evolution of player status, and are prone to causing difficulty mutations or interaction fatigue.

[0004] It is particularly worth pointing out that existing technologies have significant bottlenecks in the collaborative modeling of motion features and physiological responses. Traditional multimodal data fusion methods usually use weighted averages or simple logical splicing, and fail to establish deep correlations between motion intensity, direction, and physiological indicators. This shallow fusion results in the inability of quantum computing resources to effectively intervene in the optimization process, and motion encoding remains at the level of classical statistical features, limiting the physical interpretability and global convergence of subsequent optimization algorithms. When encountering intense motion or environmental interference, existing systems often cause misjudgments due to mismatches in motion-physiological representations, which seriously restricts the authenticity and reliability of game interactions. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent game interaction method for smart parks, which solves the problems of insufficient accuracy of action feature representation, weak multi-source data collaborative modeling capability and dynamic parameter adjustment conflicts in existing intelligent game systems.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: The intelligent game interaction method for smart park includes the following steps: Step 1: Collect the player's behavior data through multiple sensors, and encode the action features representing spatio-temporal correlation in the behavior data into quantum state representations; Step 2: Construct a dynamic Hamiltonian optimization model containing a physiological signal coupling term, and solve the quantum state representation through a quantum optimization algorithm to generate multi-objective optimization parameters; Step 3: Coordinately adjust the music rhythm, visual rendering, and task difficulty in the game interaction based on the optimization parameters; Step 4: Incrementally update the dynamic Hamiltonian optimization model according to the adjusted player feedback data to form a closed-loop optimization system.

[0007] Preferably, for the behavior data collection in Step 1: Deploy a multi-source perception network composed of a millimeter-wave radar, a visual sensor, and a biological sensor, where: ; Where: is a rotation matrix, is a translation vector, is a time offset, and are the coordinates of the k-th feature point of the radar and the visual sensor respectively.

[0008] Preferably, Step 1 includes: Collect the galvanic skin response signal GSR(t) and the blood volume pulse signal BVP(t) through a wrist-worn biological sensor, and perform time synchronization calibration with the spatio-temporal alignment data, satisfying: ; Where: : The maximum allowable time offset, : The integration window duration, represents the signal in the L2 norm within the interval, and need to be cubic spline interpolated for the sampling rate.

[0009] Preferably, the dynamic Hamiltonian optimization model is constructed as: ; The definitions of each component Hamiltonian are as follows: Participation maximization term: ; Challenge adjustment term: ; Emotion enhancement term: ; Physiological signal coupling term: ; In the formula: : Multi-objective weight coefficient, : Physiological coupling strength coefficient, GSR(t): Real-time skin conductance signal, : Pauli operator of the j-th qubit.

[0010] Preferably, the quantum optimization algorithm performs the following process: Solve the optimal parameters (γ, β) using the quantum approximate optimization algorithm: Solve the optimal parameter combination using the quantum approximate optimization algorithm , where is the phase parameter vector, is the mixing parameter vector; ; Among them, the quantum state preparation satisfies: ; In the formula: is the mixed Hamiltonian : Number of quantum circuit layers : Initial quantum state.

[0011] Preferably, the music rhythm adjustment is achieved through the following formula: ; In the formula: : Benchmark rhythm frequency, : Dynamic adjustment coefficient, : Expectation value of the participation Hamiltonian in the neutral state, : Sensitivity threshold, : Quantum expectation value of the current participation Hamiltonian.

[0012] Preferably, the visual rendering adjustment is achieved through dynamic mapping in the HSV color space: ; In the formula: : Benchmark hue angle, : Benchmark saturation, : Hue dynamic gain coefficient, : Saturation attenuation coefficient, : Variance of the quantum state in the Z direction.

[0013] Preferably, the incremental update is achieved through sliding window optimization: ; In the formula: : Hamiltonian parameter vector at time t , : learning rate, : window function with exponential decay; ; Wherein: : update period, represents the weighted average of the expected values within the window.

[0014] Preferably, the fourth step includes: Implement cross-device incremental knowledge transfer through the quantum state compression protocol: ; In the formula: : quantum coding gate operation, : optimized quantum state of the k-th qubit, : total number of qubits, migration period .

[0015] Preferably, in the incremental update process of the fourth step, real-time anomaly detection and adaptive adjustment are performed to satisfy: ; In the formula: : KL divergence of adjacent window parameter distributions, : anomaly detection threshold, : quantum state stability index, : maximum allowable learning rate.

[0016] In summary, the present invention includes at least one of the following beneficial technical effects: 1. The present invention adopts the spatio-temporal registration algorithm of millimeter-wave radar, visual sensor and biological signal, and combines the probability amplitude-phase dual-parameter mapping of the quantum state, solves the problem of the disconnection between limb movement characteristics and physiological response in traditional motion capture, and realizes the high-fidelity representation of human motion patterns in the quantum Hilbert space, providing a quantization input with clear physical meaning for subsequent optimization.

[0017] 2. The present invention globally optimizes the participation degree, challenge degree, emotional enhancement and physiological coupling terms by using the quantum approximate optimization algorithm, solves the problem of experience fragmentation caused by single-objective optimization in traditional game parameter adjustment, realizes the dynamic balance of music rhythm, visual rendering and task difficulty parameters, and ensures the continuity and adaptability of the player's flow experience.

[0018] 3. The present invention adopts a timing gradient update mechanism with an exponential decay window and combines real-time detection of the KL divergence of parameter distribution drift, solving the problem of interactive distortion caused by model overfitting and noise accumulation during long-term operation, and significantly improving the system's dynamic adaptability to sudden changes in player behavior and environmental interference.

[0019] 4. The present invention is based on a cross-device state compression transmission scheme using controlled quantum gates, solving the restriction of the scalability of the system due to the computing power limitation of a single node and data islanding, and realizing the lossless sharing of optimized knowledge among multiple devices, providing underlying technical support for the large-scale distributed game network deployment of intelligent parks. Brief Description of the Drawings

[0020] Figure 1 It is a schematic diagram of the method flow of the present invention. Detailed Embodiments

[0021] The following will Figure 1 further describe the present invention in detail in conjunction with the attached

[0022] The present invention provides an intelligent game interaction method for intelligent parks, which adopts a spatio-temporal registration algorithm for millimeter-wave radar, visual sensors and biological signals, and combines the probability amplitude-phase double-parameter mapping of quantum states to achieve a high-fidelity representation of human motion patterns in the quantum Hilbert space, providing a quantum input with clear physical meaning for subsequent optimization.

[0023] As Figure 1 shown: The intelligent game interaction method for intelligent parks may include the following steps: Step 1: Collect the behavior data of players through multiple sensors, and encode the action features representing spatio-temporal correlation in the behavior data into quantum state representations; Step 2: Construct a dynamic Hamiltonian optimization model containing a physiological signal coupling term, and solve the quantum state representation through a quantum optimization algorithm to generate multi-objective optimization parameters; Step 3: Coordinately adjust the music rhythm, visual rendering and task difficulty in the game interaction based on the optimization parameters; Step 4: Perform incremental updates on the dynamic Hamiltonian optimization model according to the adjusted player feedback data to form a closed-loop optimization system.

[0024] In this embodiment, the implementation of behavior data acquisition and fusion includes the collaborative perception and spatio-temporal alignment processing of multi-source heterogeneous sensors. By deploying a multi-modal perception network composed of millimeter-wave radars, vision sensors, and biological sensors, the limb movement characteristics and physiological response signals of players are captured in real time. Preferably, the millimeter-wave radar is configured in a high-frequency scanning mode to obtain the three-dimensional point cloud coordinates of high-precision limb joint points; the vision sensor is configured in a multi-view stereo imaging mode, and the spatial positions of joint points are extracted through a feature matching algorithm; the biological sensor is preferably a wrist-worn integrated device that synchronously collects skin conductance signals and blood volume pulse signals. The spatio-temporal alignment processing realizes multi-sensor data fusion by solving a non-linear optimization problem. For the spatio-temporal differences between the millimeter-wave radar point cloud data and the vision sensor joint point coordinates, a joint optimization objective function is constructed: ; where, is the rotation matrix, describing the spatial attitude difference between sensors; is the translation vector, representing the origin offset between sensor coordinate systems; is the time offset, used to compensate for the asynchrony of sensor sampling times; and respectively represent the three-dimensional coordinates of the th joint point detected by the radar and the vision sensor. The optimization problem is solved by the iteratively reweighted least squares method, and the fused motion trajectory after spatio-temporal alignment is output. For the synchronization of biological sensor signals and motion data, this embodiment further performs cross-modal time calibration. After the skin conductance signal GSR and the blood volume pulse signal BVP are aligned to the same sampling rate through cubic spline interpolation, the time offset is determined by minimizing the normalized signal difference: ; where, is the preset maximum allowable time offset, used to constrain the reasonable range of physiological signal transmission delay; is the integration window duration, preferably a dynamic sliding window to balance the calculation efficiency and signal stability; represents the norm of the signal in the window interval , used to eliminate the influence of individual physiological signal amplitude differences on the synchronization accuracy. The synergistic effect of spatio-temporal alignment and time calibration is that: through the rotation matrix and the translation vector , the spatial pose differences of multi-sensors are compensated, so that the radar point cloud data and the vision joint point coordinates are accurately matched in the unified coordinate system; at the same time, the time offsets and Joint optimization ensures strict synchronization of motion data and physiological signals in the time domain. The resulting multimodal fusion dataset provides input information with complete spatio-temporal correlation for subsequent quantum state encoding.

[0025] In this embodiment, the construction and solution of the dynamic Hamiltonian optimization model are achieved through quantum state encoding and multi-objective collaborative optimization. Based on the spatio-temporally aligned multimodal data obtained in Step 1, the player behavior characteristics are mapped to the quantum state space, and a Hamiltonian operator that incorporates physiological response characteristics is constructed. The optimal configuration of multi-objective parameters is solved through a quantum optimization algorithm. The quantum state encoding process includes double mapping of kinematic features and physiological signals. For the joint motion trajectories output by the fusion of millimeter-wave radar and visual sensors , extract their velocity vectors as the encoding basis for quantum bits. Preferably, the spherical coordinate mapping method is used to convert the velocity amplitude and direction angle into quantum state parameters: ; where represents the normalized velocity amplitude, reflects the motion direction angle, is the preset maximum velocity threshold. The initial quantum state is thus generated, and the quantum correlation between joints is enhanced through the global entanglement gate , where represents the entanglement connection relationship defined based on the human body bone topology. The construction of the dynamic Hamiltonian reflects the multi-objective optimization requirements, which are specifically decomposed into a weighted combination of four component operators: maximizing participation, adjusting challenge level, enhancing emotion, and physiological coupling: ; where the weight coefficient satisfies , is the physiological coupling strength coefficient. The physical meaning and technical necessity of each component Hamiltonian are as follows: Maximizing participation term: ; The kinematic activity of the th joint is reflected through the quantum expectation value , and the negative sign design makes the optimization process tend to increase the player's movement amplitude and frequency. Adjusting challenge level term: ; Based on the variance of , the player's adaptability to the level difficulty is measured. An increase in variance indicates that the player needs to adjust the strategy to cope with the challenge, thereby driving the dynamic balance of the task difficulty.

[0026] Emotional enhancement term: ; Select the square of the expected value of specific joint points (such as hands, head), amplify the quantum contribution of actions related to emotional expression, and enhance the emotional immersion of game feedback.

[0027] Physiological signal coupling term: ; Use the real-time skin conductance signal as a time-varying coupling coefficient, combine it with the operator tensor product of joint points, so that the player's physiological excitement directly modulates the quantum state evolution process and realizes physiological-behavioral cross-modal correlation. The execution of the quantum optimization algorithm adopts the framework of the quantum approximate optimization algorithm (QAOA). By alternately applying the target Hamiltonian and the mixing Hamiltonian , prepare the parameterized quantum state: ; Among them, is the initial uniform superposition state, represents the number of layers of the quantum circuit, and are the parameters to be optimized. Minimize the expected value through the classical optimizer, and output the optimal parameter combination , thereby determining the optimal configuration of the multi-objective weight coefficient and the physiological coupling strength ; Use the quantum approximate optimization algorithm to solve the optimal parameter combination , where is the phase parameter vector, is the mixing parameter vector; The phase parameter vector : Control the action duration of the target , and determine the phase accumulation process of the quantum state in the solution space. Each layer corresponds to the layer in the quantum circuit evolution time parameter, satisfying to cover the complete phase cycle.

[0028] The mixing parameter vector : Adjust the action strength of the mixing , and drive the controllable transition of the quantum state between the computational basis state and the superposition state. The parameter range is restricted to to ensure the effectiveness of the quantum gate operation.

[0029] In this embodiment, the coordinated adjustment of multi-dimensional game parameters is achieved through the dynamic mapping of quantum optimization parameters to the classical domain, specifically covering the cross-modal adaptation mechanisms of music rhythm, visual rendering, and task difficulty. Based on the multi-objective optimization parameters output in Step 2, the quantum state evolution characteristics are fused with the physiological response signals to drive the real-time tuning of game interaction elements. The dynamic adjustment of music rhythm is achieved through the non-linear mapping of the quantum participation index and the reference rhythm. According to the expected value of the participation Hamiltonian term , the hyperbolic tangent function is used to constrain the adjustment amplitude to generate the rhythm frequency adjustment formula: ; where is the preset reference rhythm frequency, reflecting the initial design intention of the game scene; is the dynamic adjustment coefficient, used to control the sensitivity of rhythm changes; represents the expected value of participation when the player is in a neutral state, calibrated through historical data statistics; is the sensitivity threshold, used to filter out small fluctuations in the quantum expected value. The introduction of the hyperbolic tangent function aims to smooth the rhythm transition and avoid auditory discomfort caused by sudden changes in quantum parameters. The real-time mapping of visual rendering adopts the HSV color space dynamic transformation mechanism, mapping the global correlation characteristics and local fluctuation characteristics of the quantum state to the hue and saturation dimensions respectively. The specific implementation is as follows: ; Hue mapping: Based on the real part value of the global quantum entanglement state , the reference hue is dynamically adjusted, where is the gain coefficient, controlling the contribution intensity of quantum correlation to hue changes. The modulo 360 operation ensures the periodicity of the hue angle, conforming to the characteristics of human visual perception. - Saturation mapping: The direction variance of the quantum state is used to characterize local quantum fluctuations, and the reference saturation is adjusted through the exponential decay function. When the quantum state tends to be stable (variance decreases), the saturation is enhanced to strengthen the visual impact; otherwise, the saturation is reduced to relieve visual fatigue. The task difficulty adaptation dynamically adjusts the target completion threshold based on the variance of the challenge degree . Define the difficulty adjustment function: ; where reflects the player's adaptation degree to the current challenge, is the adjustment intensity coefficient, is the difficulty response threshold. When , The function outputs a positive value, increasing the task objective threshold to increase the challenge; otherwise, it maintains or decreases the threshold to adapt to the player's ability level.

[0030] In this embodiment, the construction of the closed-loop optimization system is achieved through incremental model updating and distributed knowledge migration, specifically including sliding window parameter optimization, quantum state compression transmission, and anomaly adaptive adjustment mechanisms. Based on the player feedback data adjusted in step three, the Hamiltonian model parameters are dynamically corrected to ensure that the system continuously adapts to the evolution of the player's behavior pattern. Incremental parameter updating is implemented using the sliding window gradient descent algorithm to iterate the model parameters. Define the weighted expected loss function within the time window: ; where, is the window decay coefficient, controlling the decay rate of the historical data weight; is the update period, preferably matching the rhythm of the game scenario; is the window capacity. By calculating the gradient of the loss function with respect to the Hamiltonian parameter , perform parameter update: ; The technical necessity of the sliding window mechanism lies in: balancing historical experience and real-time feedback through exponentially decaying weights to avoid overfitting the latest data; the fixed update period ensures predictable allocation of computing resources to meet real-time requirements. Distributed knowledge migration is achieved through the quantum state compression protocol for cross-device collaborative optimization. For the local optimized quantum states generated by multiple player terminals, perform hierarchical compression encoding: ; where, is the preset entanglement encoding gate, compressing two qubit states into a single logical qubit; ensures that the key joint motion features are still retained in the encoded quantum state. The compressed state is broadcast to other terminals through the quantum network protocol, and the migration period and the local update period satisfy an integer multiple relationship, preferably to balance the communication overhead and the timeliness of knowledge synchronization.

[0031] Anomaly detection and adaptive adjustment are achieved through parameter distribution monitoring and quantum stability indicators. Define the KL divergence of the parameter distribution in adjacent time windows: ; When , it is determined that the system is in an abnormal state, triggering the adaptive adjustment of the learning rate: ; Among them, characterizing the quantum state directional polarization stability. The technical relevance of the adjustment mechanism lies in: the divergence monitoring parameter space mutation, the quantum stability index reflects the degree of influence of hardware noise. The two cooperate to control the decay rate of the learning rate to prevent gradient explosion or local convergence.

[0032] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent game interaction method for a smart park, characterized in that, Including the following steps: Step 1: Collect the behavior data of players through multiple sensors, and encode the action features representing spatio-temporal correlation in the behavior data into quantum state representations; Step 2: Construct a dynamic Hamiltonian optimization model including a physiological signal coupling term, and solve the quantum state representation through a quantum optimization algorithm to generate multi-objective optimization parameters; Step 3: Coordinately adjust the music rhythm, visual rendering, and task difficulty in the game interaction based on the optimization parameters; Step 4: Incrementally update the dynamic Hamiltonian optimization model according to the adjusted player feedback data to form a closed-loop optimization system.

2. The intelligent game interaction method for a smart park according to claim 1, characterized in that, The Step 1 includes: Deploy a multi-source perception network composed of a millimeter-wave radar, a visual sensor, and a biological sensor, where: ; Wherein: is the rotation matrix, is the translation vector, is the time offset, and are the coordinates of the k-th feature point of the radar and the vision sensor, respectively.

3. The intelligent game interaction method for a smart park according to claim 1, characterized in that The Step 1 includes: Collect the galvanic skin response signal GSR(t) and the blood volume pulse signal BVP(t) through a wrist-worn biological sensor, and perform time synchronization calibration for the spatio-temporally aligned data, satisfying: ； Wherein: : The maximum allowable time offset, : The integration window duration, represents the L2 norm of the signal within the interval, and needs to perform cubic spline interpolation on the sampling rate.

4. The intelligent game interaction method for a smart park according to claim 1, wherein The dynamic Hamiltonian optimization model is constructed as: ； Where the definitions of each component Hamiltonian are as follows: Participation maximization term: ； Challenge adjustment term: ； Emotion enhancement term: ； Physiological signal coupling term: ； Where: : Multi-objective weight coefficient, : Physiological coupling strength coefficient, GSR(t): Real-time skin conductance signal, : Pauli operator of the j-th qubit.

5. The intelligent game interaction method for a smart park according to claim 1, characterized in that, The quantum optimization algorithm performs the following process: Use the quantum approximate optimization algorithm to solve the optimal parameters (γ,β): Using the quantum approximate optimization algorithm to solve for the optimal parameter combination , where is the phase parameter vector, is the hybrid parameter vector; ； Where the quantum state preparation satisfies: ； Wherein: is the hybrid Hamiltonian : the number of quantum circuit layers : the initial quantum state.

6. The intelligent game interaction method for a smart park according to claim 1, characterized in that The music rhythm adjustment is achieved through the following formula: ； Wherein: : reference rhythm frequency, : dynamic adjustment coefficient, : expected value of the Hamiltonian of the participation in the neutral state, : sensitivity threshold, : quantum expected value of the current Hamiltonian of the participation.

7. The intelligent game interaction method for a smart park according to claim 1, wherein The visual rendering adjustment is achieved through dynamic mapping in the HSV color space: ； In the formula: : reference hue angle, : reference saturation, : hue dynamic gain coefficient, : saturation attenuation coefficient, : variance of the quantum state in the Z direction.

8. The intelligent game interaction method for a smart park according to claim 1, characterized in that, The incremental update is achieved through sliding window optimization: ； where: : Hamiltonian parameter vector at time t , : learning rate, : window function with exponential decay; ； Wherein: : Update period Indicates the weighted average of the expected values within the window.

9. The intelligent game interaction method for a smart park according to claim 1, wherein The Step 4 includes: Achieve cross-device incremental knowledge transfer through a quantum state compression protocol: ； In the formula: : Quantum coding gate operation, : Optimized quantum state of the k-th qubit, : Total number of qubits, migration period .

10. The intelligent game interaction method for a smart park according to claim 1, wherein In the incremental update process in Step 4, real-time anomaly detection and adaptive adjustment are performed, satisfying: ； In the formula: : KL divergence of adjacent window parameter distribution : Anomaly detection threshold : Quantum state stability index : Maximum allowable learning rate

Images