Doll having intelligent simulated eyes

Abstract

The present invention relates to the technical field of dolls, and in particular to a doll having intelligent simulated eyes, comprising a support body, wherein the support body is provided with simulated eyes and an inflation pump, the simulated eyes being LCD screens, and the inflation pump being electrically connected to the LCD screens; and an inflation cavity is provided inside the support body, an output end of the inflation pump being in communication with the inflation cavity. In the present invention, by means of the arrangement of the inflation pump and the LCD screens on the support body, the inflation pump and the LCD screens being electrically connected to each other, the linkage between the inflation state of the inflation pump and the display state of the LCD screens can be realized, thereby giving the doll a novel, vivid and dynamic effect.

Description

A doll with intelligent lifelike eyes Technical Field

[0001] This invention relates to the field of doll technology, specifically to a doll with intelligent simulated eyes. Background Technology

[0002] In today's amusement and commercial settings, inflatable air dome structures are widely used as a common facility. However, most inflatable air dome structures on the market suffer from limited functionality and rigid forms. Traditional air dome structures are merely static objects, providing only basic spatial division or aesthetic display functions, failing to fully engage users and stimulate their interest. For example, in amusement parks and theme parks, air dome structures are often just simple architectural shapes, lacking interactivity with visitors. During festivals or special events, their unchanging appearance makes it difficult to create a matching lively atmosphere, failing to meet people's pursuit of immersive and fun experiences. Furthermore, in daily use, these rigid air dome structures cannot be flexibly adjusted according to different scenarios and needs, making their role in enhancing the recreational atmosphere extremely limited. Technical issues

[0003] Lacking dynamic functional design, air-supported membrane structures fail to exhibit fluid changes, making it difficult to attract tourists to linger and participate for extended periods. Furthermore, their monotonous appearance fails to reflect the unique themes of different festivals and events, resulting in insufficient atmosphere and significantly limiting the potential of air-supported membrane products to enhance user experience and commercial value. Therefore, developing a new type of inflatable air-supported membrane structure with dynamic functions and a vibrant feel, thereby significantly enhancing the recreational atmosphere and enriching usage scenarios, is an urgent practical need and of great significance. Technical solutions

[0004] The purpose of this invention is to address the aforementioned shortcomings of the prior art by providing a doll with intelligent, lifelike eyes.

[0005] The objective of this invention is achieved through the following technical solution: a doll with intelligent simulated eyes, comprising a support body; the support body is provided with simulated eyes and an air pump; the simulated eyes are an LCD screen; the air pump is electrically connected to the LCD screen; the support body is provided with an inflation cavity; the output end of the air pump is connected to the inflation cavity.

[0006] The present invention is further configured such that a PVC sheet is provided between the LCD screen and the supporting body.

[0007] The present invention is further configured such that the PVC sheet extends to provide a connecting seat; the connecting seat is detachably connected to a waterproof back cover; a receiving cavity is formed between the connecting seat and the waterproof back cover; the LCD screen is disposed in the receiving cavity; and the PVC sheet is provided with a display channel communicating with the receiving cavity.

[0008] The present invention is further configured such that a waterproof silicone sheet is provided between the LCD screen and the connector.

[0009] The present invention is further configured such that the PVC sheet is connected to the support body by adhesive and / or sewing.

[0010] The present invention is further configured such that the doll with intelligent simulated eyes also includes a rotating disk; the supporting body is disposed on the rotating disk.

[0011] The present invention is further configured to include the following steps:

[0012] S101. Obtain real-time data of the air pump during operation to obtain dynamic correlation data between load fluctuation and power supply distribution;

[0013] S102. Determine the output status of the air pump based on the dynamic correlation data, and adjust the voltage output range of the air pump.

[0014] S103. Determine the power supply stability of the LCD screen based on the adjusted voltage output range, and use a voltage regulator module to compensate and adjust the power supply line.

[0015] S104. Adjust the mounting structure of the LCD screen according to the pressure changes inside the inflation chamber;

[0016] S105. Determine the stability of power distribution based on the stress data of the adjusted installation structure, and generate the optimal power distribution model.

[0017] The present invention is further configured such that acquiring real-time data of the air pump operation to obtain dynamic correlation data between load fluctuations and power supply distribution includes:

[0018] The current, air pressure, and LCD screen power supply voltage values ​​of the air pump during operation are acquired by sensors. The digital data is processed by time series analysis to determine the frequency components and dynamic correlation strength of load fluctuations. The influence coefficient of voltage stability on power distribution is determined by regression analysis, and the classification model results are generated by support vector machine algorithm.

[0019] The present invention is further configured such that, determining the output status of the air pump based on the dynamic correlation data and adjusting the voltage output range of the air pump includes:

[0020] The output value is extracted from the dynamic association dataset and compared with the preset threshold. If it exceeds the preset range, the target voltage output range value is obtained, the voltage parameters of the air pump drive module are adjusted, the control parameters are optimized using the random forest algorithm, and stable output state data is generated.

[0021] The present invention is further configured such that, in determining the power supply stability of the LCD screen based on the adjusted voltage output range, and using a voltage regulator module to compensate and adjust the power supply line, the following steps are taken:

[0022] The adjusted voltage output range is compared with the power supply threshold of the LCD screen to extract the characteristics of the unstable state. The adjustment parameters of the voltage regulator module are obtained from the preset module. The deviation characteristics are optimized by the support vector machine algorithm, the control logic of the voltage regulator module is updated, and the stable voltage output data is generated.

[0023] The present invention is further configured such that the mounting structure for adjusting the LCD screen according to the pressure change inside the inflation cavity includes:

[0024] Real-time pressure data inside the inflation chamber is obtained, and the time series data of pressure change is processed by a convolutional neural network to determine the spatial compression quantification index. The deformation is predicted by a regression algorithm, and the stress distribution is simulated by finite element analysis to verify the stability of the adjustment scheme and obtain the optimized installation structure parameters.

[0025] The step of determining the stability of power distribution based on the adjusted installation structure stress data and generating an optimal power distribution model includes:

[0026] The pressure change trend and stress data inside the inflation chamber are analyzed to determine the fluctuation conditions and threshold range. Historical data are classified and trained using a support vector machine, and hyperparameters are adjusted to generate a classification model. Logistic regression is used to analyze historical power allocation data to generate the optimal power allocation model and output the optimization results. Beneficial effects

[0027] The beneficial effects of the present invention are as follows: By setting an air pump and an LCD screen on the support body and electrically connecting the air pump and the LCD screen, the air pump inflation status and the LCD screen display status can be linked, thereby giving the doll a new effect of being spiritual and lively. Attached Figure Description

[0028] The invention will be further illustrated with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the invention. For those skilled in the art, other drawings can be obtained based on the following drawings without any creative effort.

[0029] Figure 1 is a schematic diagram of the structure of the present invention;

[0030] Figure 2 is a schematic diagram of the structure of the present invention, which combines the supporting body with the simulated eye.

[0031] Figure 3 is a cross-sectional view of the support body of the present invention in conjunction with the simulated eye;

[0032] Figure 4 is an exploded view of the structure of the present invention, showing the cooperation between the supporting body and the simulated eye.

[0033] Figure 5 is a flowchart of the present invention;

[0034] The components include: 1. Support body; 2. Simulated eyes; 3. Air pump; 4. PVC sheet; 41. Connecting seat; 42. Display channel; 5. Waterproof back cover; 51. Receiving cavity; 6. Waterproof silicone sheet; 7. LCD screen; 8. Rotating disk. Embodiments of the present invention

[0035] The present invention will be further described in conjunction with the following embodiments.

[0036] As shown in Figures 1 to 5, the doll with intelligent simulated eyes 2 described in this embodiment includes a support body 1; the support body 1 is provided with simulated eyes 2 and an air pump 3; the simulated eyes 2 are LCD screens 7; the air pump 3 is electrically connected to the LCD screen 7; the support body 1 is provided with an inflation cavity; the output end of the air pump 3 is connected to the inflation cavity.

[0037] Specifically, in this embodiment, by setting an air pump 3 and an LCD screen 7 on the support body 1, and electrically connecting the air pump 3 and the LCD screen 7, the inflation state of the air pump 3 and the display state of the LCD screen 7 can be linked, thereby giving the doll a new effect of being spiritual and lively.

[0038] The LCD screen 7 can be connected via Bluetooth or directly import multiple effect videos, and has the function of switching between effect videos; the circuit board control design and modules of the LCD screen 7 can be synchronized with the air pump 3.

[0039] In this embodiment, a doll with intelligent simulated eyes 2 is provided, wherein a PVC sheet 4 is provided between the LCD screen 7 and the supporting body 1. This arrangement facilitates the connection and fixation of the LCD screen 7 and the supporting body 1.

[0040] In this embodiment, a doll with intelligent simulated eyes 2 is described. The PVC sheet 4 extends to form a connecting base 41. A waterproof back cover 5 is detachably connected to the connecting base 41. A receiving cavity 51 is formed between the connecting base 41 and the waterproof back cover 5. An LCD screen 7 is disposed within the receiving cavity 51. A display channel 42 communicating with the receiving cavity 51 is provided through the PVC sheet 4. In this embodiment, a waterproof silicone sheet 6 is provided between the LCD screen 7 and the connecting base 41. This design provides waterproofing. In this embodiment, the PVC sheet 4 is connected to the support body 1 by glue and / or sewing.

[0041] Specifically, firstly, the PVC sheet 4 and the fabric support body 1 are glued together with an oil-based glue. Then, colored thread of the same color as the PVC sheet 4 is used to sew the PVC sheet 4 and the fabric support body 1 together. A waterproof silicone sheet 6 is added between the PVC sheet 4 and the fabric support body 1 to provide waterproofing and reinforcement, prevent it from falling off during use, and ensure quality and safe use.

[0042] After the LCD screen is assembled, it is encapsulated with full-adhesive glue to provide waterproofing and abrasion protection.

[0043] Then place the LCD screen in the middle of the connector 41, and then use screws to fix the connector 41 to the waterproof back cover 5.

[0044] Additionally, glue is applied between the connector 41 and the waterproof back cover 5 to prevent water ingress and detachment. Furthermore, an arc-shaped pocket is sewn at the nearby fabric cover seam and sealed with Velcro to prevent the connector 41 from falling off, thus providing protection.

[0045] This embodiment describes a doll with intelligent simulated eyes 2, which also includes a rotating disk 8; the support body 1 is disposed on the rotating disk 8. This arrangement enables the support body 1 to rotate.

[0046] The doll with intelligent simulated eyes 2 described in this embodiment includes the following steps:

[0047] S101. Obtain real-time data of the air pump 3 during operation to obtain dynamic correlation data between load fluctuation and power supply distribution;

[0048] S102. Determine the output status of the air pump 3 based on the dynamic correlation data, and adjust the voltage output range of the air pump 3.

[0049] S103. Determine the power supply stability of LCD screen 7 based on the adjusted voltage output range, and use a voltage regulator module to compensate and adjust the power supply line.

[0050] S104. Adjust the mounting structure of LCD screen 7 according to the pressure change inside the inflation chamber;

[0051] S105. Determine the stability of power distribution based on the stress data of the adjusted installation structure, and generate the optimal power distribution model.

[0052] Specifically, during the research and development process, a challenge arose of ensuring efficient inflation of the inflatable cavity while maintaining stable operation of the LCD screen 7. Both depend on the coordinated operation of the air pump 3. In detail, the air pump 3, as the core power component, has its output directly connected to the inflatable cavity within the support body 1, responsible for injecting gas to form the doll's shape. However, the air pump 3 also needs to establish an electrical connection with the LCD screen 7 to drive the display function of the simulated eyes 2. This leads to a unique business scenario challenge: during the doll's transformation from its initial flat state to full inflation, the workload of the air pump 3 dynamically changes, and this load fluctuation can interfere with its power output, thus affecting the display stability of the LCD screen 7. For example, when the internal pressure of the inflatable cavity rises rapidly, the air pump 3 may prioritize airflow output, resulting in insufficient voltage or current supply to the LCD screen 7, causing the simulated eyes 2 to flicker, go black, or experience response delays. Conversely, if the power needs of the LCD screen 7 are prioritized, inflation efficiency may decrease, prolonging the doll's forming time and impacting the user experience. Furthermore, the limited space within the support body 1 means that the expansion of the air chamber may compress the mounting position of the LCD screen 7, causing the screen to deform under stress or experience poor contact, further complicating the technical implementation. The core of this contradiction lies in the fact that while the air pump 3 and the LCD screen 7 are functionally related, their operating rhythms, resource requirements, and spatial layouts are mismatched, creating a deep-seated technical problem that urgently needs to be solved during the research and development process.

[0053] This embodiment establishes a dynamic correlation model between load fluctuations and power distribution by real-time monitoring of the current and air pressure of the air pump 3 and the power supply voltage of the LCD screen 7. When the output of the air pump 3 exceeds the stable range, the voltage output is automatically adjusted, and the voltage stability risk of the LCD screen 7 is assessed. If necessary, compensation is performed through a voltage stabilization module. The impact of pressure changes within the inflation chamber on the installation position of the LCD screen 7 is also considered, and the installation structure is optimized based on the degree of compression. If the internal pressure affects the stability of power distribution, a support vector machine is used to classify and train historical data to generate an optimal power distribution model. This method achieves coordinated control of the air pump 3 and the LCD screen 7, improving the overall stability and reliability of the system and providing a new approach for collaborative equipment management in similar scenarios.

[0054] The doll with intelligent simulated eyes 2 described in this embodiment, wherein the acquisition of real-time data of the air pump 3 during operation, and obtaining dynamic correlation data between load fluctuations and power supply distribution, includes:

[0055] The current value, air pressure value and LCD screen 7 power supply voltage value of the air pump 3 during operation are obtained by the sensor. The digital data is processed by time series analysis to determine the frequency components and dynamic correlation strength of load fluctuations. The influence coefficient of voltage stability on power supply distribution is determined by regression analysis. The classification model results are generated by support vector machine algorithm.

[0056] Specifically, the process includes the following steps: Step 1: Acquire real-time current and air pressure values ​​of the air pump 3 and the power supply voltage of the LCD screen 7 during operation using sensors, and obtain digitized data through analog-to-digital conversion. Step 2: Process the digitized data using time series analysis to determine the characteristic parameters of real-time current changes, air pressure changes, and voltage stability. Step 3: If the real-time current change exceeds a preset threshold, decompose the frequency components of the load fluctuation using Fourier transform to obtain frequency domain characteristic data. Step 4: Based on the correspondence between the frequency domain characteristic data and air pressure changes, use correlation analysis to determine the dynamic correlation strength between load fluctuations and air pressure values. Step 5: Process the power supply voltage and power distribution data through regression analysis to determine the influence coefficient of voltage stability on power distribution. Step 6: Obtain the dynamic correlation strength and influence coefficient, and use the support vector machine algorithm to classify the correlation between load fluctuations and power distribution to obtain the classification model results. Step 7: Compare the classification model results with the real-time acquired data to determine the dynamic correlation trend between load fluctuations and power distribution, and output trend data.

[0057] Acquiring real-time current and air pressure values ​​of the air pump 3 and the power supply voltage of the LCD screen 7 via sensors, and converting them into digital data, forms the foundation of the entire analysis process. As one implementation, a current sensor can be used to monitor the operating current of the air pump 3 motor in real time, for example, stabilizing at around 2.5A during normal operation. The air pressure sensor records the air pump's output air pressure, for example, maintaining a standard value of 3 bar. Furthermore, a voltage sensor can detect the power supply voltage of the LCD screen 7, for example, a stable 5V. This data is converted into digital signals by an analog-to-digital converter at a sampling rate of 100 times per second, providing high-precision input for subsequent analysis. The advantage of this method is its ability to capture subtle changes during operation, laying the foundation for anomaly detection. Time series analysis is used to process this digital data, enabling the extraction of key characteristic parameters. Specifically, the trend of current changes over time can be observed; for example, at the initial startup of the air pump 3, the current may briefly rise to 3A and then fall back to 2.5A, indicating the characteristics of the motor starting load. The air pressure value may gradually increase from 0 bar to 3 bar, reflecting the dynamics of the inflation process. Voltage stability is manifested by power supply voltage fluctuations not exceeding ±0.1V. This analysis helps identify operational patterns, such as determining whether equipment is in normal condition through the stationarity of time series data, thereby improving the accuracy of fault early warning. When real-time current changes exceed a preset threshold, such as 3.5A, the frequency components of load fluctuations can be decomposed using Fourier transform. For example, in a certain operation, if current fluctuations are frequent and large in amplitude, the Fourier transform results show that the main frequencies are concentrated at 10Hz and 50Hz, which may correspond to abnormal motor speed or load changes caused by air circuit blockage. This frequency domain characteristic data can intuitively reflect the source of fluctuations and pinpoint the root cause of the problem more accurately than time domain analysis, which helps to optimize equipment maintenance strategies. Furthermore, based on the correspondence between frequency domain characteristic data and air pressure changes, correlation analysis can be used to determine the dynamic correlation strength between load fluctuations and air pressure values. For example, assuming that a 10Hz frequency fluctuation is highly correlated with an air pressure drop to 2.5 bar, with a correlation coefficient of 0.9, it indicates that the load fluctuation may be caused by insufficient air pressure. The advantage of this analysis is that it reveals the causal relationship between variables, providing a basis for subsequent classification. It should be noted that abnormal air pressure values ​​may indicate air leakage or decreased pump efficiency; correlation strength can quickly pinpoint the problem. By processing the supply voltage and power distribution data through regression analysis, the impact coefficient of voltage stability can be determined. For example, assuming the LCD screen flickers when the voltage drops from 5V to 4.8V, regression analysis shows that for every 0.1V drop in voltage, the power distribution efficiency decreases by 5%, with an impact coefficient of -0.05. This method quantifies the impact of voltage fluctuations on the system, providing data support for power supply design optimization and improving overall stability. After obtaining the dynamic correlation strength and impact coefficient, a support vector machine algorithm can be used to classify the correlation between load fluctuations and power distribution.For example, data can be categorized into "normal operation" and "abnormal fluctuations." When training the model, a correlation strength of 0.9 and an influence coefficient of -0.05 are input. The results show that a particular operation is classified as "abnormal fluctuation." The advantage of this classification model is its ability to automatically identify complex patterns, reduce the subjectivity of manual judgment, and improve the efficiency of real-time monitoring. By comparing the classification model results with real-time collected data, the dynamic correlation trend between load fluctuations and power distribution can be determined. For example, real-time data shows a current fluctuation frequency of 50Hz, a pressure drop to 2.5 bar, and a voltage stabilization at 5V. The model predicts "abnormal fluctuation," and the trend data suggests that abnormal pressure may be the main cause. The benefit of this comparative analysis is the ability to dynamically adjust operating strategies, such as reducing pump speed to alleviate insufficient pressure and extend equipment lifespan. For example, as an extension, threshold settings can be optimized by combining historical data. If the current exceeds 3.5A and pressure drops during multiple operations, the threshold can be dynamically adjusted to 3.3A for early warning. Furthermore, if voltage fluctuations are frequent, a voltage stabilization module can be added to ensure stable power distribution. This multi-faceted approach not only enhances the rigor of the analysis but also provides direction for equipment optimization through trend data, ultimately achieving a dual improvement in operational efficiency and reliability.

[0058] The doll with intelligent simulated eyes 2 described in this embodiment, wherein the step of determining the output state of the air pump 3 based on the dynamic correlation data and adjusting the voltage output range of the air pump 3 includes:

[0059] The output value is extracted from the dynamic association dataset and compared with the preset threshold. If it exceeds the preset range, the target voltage output range value is obtained, the voltage parameters of the air pump 3 drive module are adjusted, the control parameters are optimized using the random forest algorithm, and stable output state data is generated.

[0060] Specifically, the process includes the following steps: Step 1: Collect real-time operating data of the air pump 3 using sensors to generate a dynamic correlation dataset, obtaining a preliminary output state. Step 2: Extract the output value of the air pump 3 from the dynamic correlation dataset and compare it with a preset stable range threshold to determine if the output exceeds the threshold. Step 3: If the comparison determines that the output of the air pump 3 exceeds the preset stable range, obtain the target voltage output range value according to a predetermined interval table. Step 4: Adjust the voltage parameters of the air pump 3 drive module based on the obtained target voltage output range value to obtain adjusted output state data. Step 5: Perform a secondary matching analysis between the adjusted output state data and the correlation data to determine whether the predetermined interval requirement has been met. Step 6: If the secondary matching analysis determines that the predetermined interval has not been met, use a random forest algorithm to optimize the features of the dynamic correlation dataset to obtain optimized control parameters. Step 7: Update the voltage output range of the air pump 3 according to the optimized control parameters to generate the final stable output state data.

[0061] Real-time data collection during the operation of the air pump 3 via sensors forms the basis for generating a dynamic correlation dataset. As one implementation method, current sensors, pressure sensors, and voltage sensors can be used to collect the operating current of the air pump 3 motor, the output air pressure, and the power supply voltage of the drive module, respectively. For example, during a certain operation, the current sensor records a brief increase in current from 2.5A to 3A when the motor starts, the pressure sensor shows the air pressure gradually increasing from 0 bar to 3 bar, and the voltage sensor detects that the power supply voltage stabilizes at 5V. This data is converted into digital signals at a high sampling rate using an analog-to-digital converter, forming a dynamic dataset containing timestamps. The advantage of this method is its ability to comprehensively capture the real-time status of the equipment, providing a reliable basis for subsequent analysis. Specifically, by extracting output values ​​from the dynamic correlation dataset and comparing them with preset thresholds, it can be determined whether the equipment deviates from its normal operating range. For example, the preset current stability range is 2.3A to 2.7A, and the standard air pressure value is 3 bar ± 0.2 bar. If a certain data collection shows that the current reaches 3.2A and the air pressure drops to 2.6 bar, it exceeds the preset range. This comparison not only quickly locates anomalies but also provides direction for adjustment strategies. It should be noted that the threshold setting can be optimized based on historical operating data. For example, by determining through multiple tests that the equipment efficiency decreases when the current exceeds 2.8A, the accuracy of the judgment can be improved. Furthermore, when the output exceeds the stable range, the voltage can be adjusted according to a predetermined range table. For example, the range table specifies that when the air pressure is below 2.8 bar, the target voltage range is 5.2V to 5.5V. Assuming the air pressure is 2.6 bar, adjusting the drive module voltage from 5V to 5.3V will increase the motor power, and the air pressure will gradually recover to 3 bar. The advantage of this adjustment is that it optimizes the operating state through voltage parameters, avoiding overload or efficiency loss due to load fluctuations. The adjusted output status data needs to be matched and analyzed with dynamically correlated data to verify whether the expected results have been achieved. For example, if the current stabilizes at 2.6A after adjustment, the air pressure recovers to 3 bar, and the voltage is 5.3V, matching the predetermined range, the adjustment is effective. This secondary verification ensures the reliability of the adjustment strategy and reduces the risk of misadjustment. As one implementation method, time series graphs can be drawn using visualization tools to visually display the parameter change trends before and after adjustment, improving the transparency of the analysis. If the secondary matching fails to meet expectations, a random forest algorithm is needed to optimize the features. For example, in one operation, the air pressure remained at 2.7 bar after adjustment, failing to reach 3 bar. Random forest analysis of historical data revealed a strong correlation between current fluctuation frequency (10Hz) and insufficient air pressure, while voltage stability had a relatively small impact on air pressure. Therefore, control parameters were optimized, focusing on adjusting the motor speed rather than the voltage. This method generates a more precise control strategy by mining deep patterns in the data, improving adjustment efficiency. Updating the voltage output range based on the optimized parameters is crucial for achieving final stability.Specifically, if the random forest suggests reducing the rotation speed and maintaining a voltage of 5.1V, the updated data shows a current of 2.5A, with the air pressure stabilizing at 3 bar and reduced fluctuations. This update not only resolves the current anomaly but also provides a reference for subsequent operations. Furthermore, thresholds can be dynamically adjusted based on historical data, such as optimizing the current limit from 2.8A to 2.7A to provide early warnings of potential risks. As an extended solution, trend analysis can be introduced to optimize the overall strategy. For example, if multiple runs reveal a correlation between air pressure drops and ambient temperature, adding a temperature sensor to collect data and adjusting the model input can improve classification accuracy. This multi-faceted approach not only enhances the rigor of the solution but also optimizes equipment lifespan and operating efficiency through data-driven optimization. For example, adding a voltage regulator module when voltage fluctuations are frequent ensures stable power supply and reduces LCD screen flickering. Ultimately, these measures collectively improve the reliability and performance of the air pump 3, providing solid support for real-time monitoring and maintenance.

[0062] This embodiment describes a doll with intelligent simulated eyes 2. The step of determining the power supply stability of the LCD screen 7 based on the adjusted voltage output range and using a voltage regulator module to compensate and adjust the power supply line includes:

[0063] The adjusted voltage output range is compared with the power supply requirement threshold of LCD screen 7. The characteristics of the unstable state are extracted, the adjustment parameters of the voltage regulator module are obtained from the preset module, the deviation characteristics are optimized by the support vector machine algorithm, the control logic of the voltage regulator module is updated, and the stable voltage output data is generated.

[0064] Specifically, the adjusted voltage output data is compared in real time with the power supply demand threshold of the LCD screen 7 to determine whether the voltage output exceeds the stable range, thus obtaining a preliminary judgment result. Based on the preliminary judgment result, features of the unstable state are extracted. If instability exists, the initial adjustment parameters of the voltage regulator module are obtained from the preset modules to determine the compensation adjustment direction. For the obtained compensation adjustment direction, the voltage parameters of the power supply line are dynamically adjusted by the voltage regulator module to obtain adjusted power supply line data. The voltage output value is extracted from the adjusted power supply line data and subjected to a secondary matching analysis with the power supply demand threshold of the LCD screen 7 to determine whether the stable state requirement has been met. If the secondary matching analysis determines that the stable state requirement has not been met, a support vector machine algorithm is used to optimize the deviation features between the voltage output and the power supply demand, obtaining optimized adjustment parameters. The control logic of the voltage regulator module is updated based on the optimized adjustment parameters to implement precise compensation adjustment of the power supply line, generating stable voltage output data. The stable voltage output data is correlated with the operating status of the LCD screen 7 for verification, and a decision tree algorithm is used to analyze the matching degree between the voltage output and the power supply demand to determine the final adjustment result.

[0065] Specifically, by collecting and adjusting the voltage output data and comparing it in real time with the power supply demand threshold of the LCD screen 7, it is determined whether the voltage output exceeds the stable range, thus obtaining a preliminary judgment result. The core of this process lies in real-time monitoring and threshold comparison. As one implementation method, a voltage sensor can collect data 10 times per second. Assuming the power supply demand threshold of the LCD screen 7 is 5V±0.2V, if a voltage value of 5.3V is collected in a certain instance, it exceeds the stable range and is initially judged as unstable. This real-time nature ensures that the system can quickly respond to abnormal fluctuations. For example, during the testing of the LCD screen 7 on the production line, if the voltage suddenly jumps to 5.4V, the system can immediately mark it as abnormal, reducing the risk of subsequent misoperation. Based on the preliminary judgment result, the characteristics of the unstable state are extracted. If instability exists, the initial adjustment parameters of the voltage regulator module are obtained from the preset module to determine the direction of compensation adjustment. Specifically, the characteristics of the unstable state may manifest as the frequency and amplitude of voltage fluctuations. For example, if the voltage repeatedly jumps between 5.1V and 5.3V within 5 seconds, feature extraction can identify a fluctuation amplitude of 0.2V and a frequency of 2 times per second. The initial adjustment parameters stored in the preset module may be empirical values ​​derived from historical data, such as setting the output gain of the voltage regulator module to +0.1V with upward compensation. This approach allows for rapid problem identification and provides preliminary solutions, preventing continuous voltage imbalances that could cause flickering or damage to the LCD screen 7. Based on the acquired compensation adjustment direction, the voltage regulator module dynamically adjusts the voltage parameters of the power supply line to obtain adjusted power supply line data. Furthermore, the voltage regulator module can employ a PID control strategy to adjust the output in real time according to the deviation. For example, with an initial voltage of 5.3V and a target of 5V, the voltage regulator module gradually reduces the output power to adjust the voltage to 5.1V. The adjusted power supply line data may show a voltage stable between 5.05V and 5.1V for 10 consecutive seconds. This dynamic adjustment improves the system's adaptability, effectively maintaining power supply stability, especially when external loads change. The voltage output value is extracted from the adjusted power supply line data and subjected to a secondary matching analysis with the LCD screen 7's power supply demand threshold to determine whether a stable state has been achieved. As one implementation method, 10 sets of data are extracted, with an average value of 5.08V, all within the range of 5V ± 0.2V, which is considered to have reached a stable state. If any value is 5.25V, the secondary matching fails. This secondary verification improves the accuracy of the judgment. For example, in high-brightness mode, the LCD screen 7 may be more sensitive to voltage, and secondary matching can ensure that the display effect is not affected by small deviations. If the secondary matching analysis determines that the stable state requirement has not been met, a support vector machine algorithm is used to optimize the deviation characteristics between voltage output and power supply demand, obtaining optimized adjustment parameters. Specifically, the support vector machine can analyze the time series data of voltage deviation to identify key influencing factors, such as voltage drift caused by temperature rise.Assuming the analysis reveals a positive correlation between the deviation and ambient temperature, the optimized parameters might be adjusted to reduce the compensation by 0.05V. The advantage of this machine learning method lies in its ability to uncover hidden patterns and enhance the intelligence of regulation. For example, in a high-temperature workshop during summer, optimized parameters can significantly reduce overvoltage. The control logic of the voltage regulator module is updated based on the optimized adjustment parameters, implementing precise compensation regulation of the power supply line to generate stable voltage output data. Furthermore, the control logic can be set to update parameters every minute; if the optimized parameter is a reduction of 0.05V, the voltage will be adjusted from 5.25V to 5V, with stable output data showing fluctuations of no more than 0.1V within 30 seconds. This precise compensation ensures long-term operational reliability. For example, in a 24-hour continuous operation LCD screen 7 testing device, it effectively avoids malfunctions caused by accumulated voltage deviations. By correlating the stable voltage output data with the operating status of the LCD screen 7, a decision tree algorithm is used to analyze the matching degree between the voltage output and power supply requirements, determining the final adjustment result. As one implementation method, the decision tree can determine the matching degree based on inputs such as voltage value, current load, and LCD screen 7 brightness. For example, when the voltage is 5V and the current is 0.5A, the brightness is normal, and the matching degree is 95%; if the voltage is 5.3V, the brightness is abnormal, and the matching degree drops to 70%. The advantage of this analysis method is its intuitiveness and strong interpretability, providing a basis for subsequent optimization. It should be noted that a high matching degree can reduce screen flicker or color difference issues, improving the user experience. For example, in medical display devices, a high matching degree can also ensure the accuracy of diagnostic images. The above solution, from real-time monitoring to intelligent optimization, forms a complete voltage stability control logic. Through multi-level verification and algorithm support, it ensures the stability of the LCD screen's power supply and improves the system's adaptive capability. The practical effect of this method is particularly evident in industrial scenarios, significantly reducing equipment failure rates and extending service life.

[0066] The doll with intelligent simulated eyes 2 described in this embodiment includes a mounting structure for adjusting the LCD screen 7 according to changes in the internal pressure of the inflatable cavity, comprising:

[0067] Real-time pressure data inside the inflation chamber is obtained, and the time series data of pressure change is processed by a convolutional neural network to determine the spatial compression quantification index. The deformation is predicted by a regression algorithm, and the stress distribution is simulated by finite element analysis to verify the stability of the adjustment scheme and obtain the optimized installation structure parameters.

[0068] Real-time pressure data inside the inflation chamber collected by sensors is acquired. The amplitude and frequency of pressure changes are calculated within a preset time window to obtain pressure change characteristic values. Dynamic parameters affecting the installation position are extracted from these characteristic values. The time-series data of pressure changes is processed using a convolutional neural network to determine the quantitative index of spatial compression. The compression degree is calculated based on the quantitative index. If the compression degree exceeds a preset threshold, a regression algorithm is used to predict the deformation at the installation position, obtaining deformation distribution data. The deformation distribution data is acquired, and the distribution of stress points at the LCD screen 7 installation position is analyzed using a pre-established geometric model of the installation structure to determine the adjustment direction of the installation structure. A parameter adjustment scheme is generated based on the adjustment direction. The boundary conditions and material properties of the installation structure are processed using an iterative optimization algorithm to obtain the adjusted structural parameter set. Key geometric features are extracted from the adjusted structural parameter set, and the stress distribution under pressure changes in the inflation chamber is simulated using finite element analysis to determine the stability of the adjustment scheme. The stability judgment results are obtained, and the adjustment scheme is verified to meet the compression degree constraint using preset convergence conditions to obtain the final optimized installation structure parameters.

[0069] Specifically, acquiring real-time pressure data inside the inflation chamber collected by sensors is fundamental to optimizing the installation structure. As one implementation method, a high-sensitivity pressure sensor can be used to continuously collect pressure data for 30 seconds at a sampling frequency of 100Hz. Specifically, by setting a 1-second time window, the amplitude and frequency of pressure changes are calculated to obtain pressure change characteristic values. For example, if the pressure changes from 100kPa to 110kPa in 1 second, the amplitude is 10kPa and the frequency is 1Hz. Further, dynamic parameters affecting the installation location are extracted from the pressure change characteristic values. This can be achieved by processing the time-series data of pressure changes using a convolutional neural network. Specifically, a network structure containing 3 convolutional layers and 2 fully connected layers is used, with 30 seconds of pressure data as input and a quantitative index representing the degree of spatial compression as output. For example, the network might output a value between 0 and 1, with 0.8 indicating a high degree of compression. It should be noted that when calculating the compression degree value based on the quantitative index of spatial compression, a threshold, such as 0.7, can be set. When the compression degree value exceeds this threshold, a regression algorithm is needed to predict the deformation of the installation location. One implementation method is to use the Support Vector Regression (SVR) algorithm, taking pressure change characteristics and compression degree values ​​as inputs and predicting deformation at each point of the installation location as outputs. After obtaining the deformation distribution data, the stress distribution at the installation location of the LCD screen 7 is analyzed using a pre-established geometric model of the installation structure. Specifically, a three-dimensional model containing the LCD screen 7, support structure, and inflatable cavity can be built using finite element analysis software. By applying the predicted deformation, the stress distribution at each location is analyzed to determine the stress concentration area, thereby determining the adjustment direction of the installation structure. When generating a parameter adjustment scheme based on the adjustment direction, an iterative optimization algorithm can be used to handle the boundary conditions and material properties of the installation structure. For example, a genetic algorithm can be used to optimize the geometric parameters and material selection of the support structure to minimize stress concentration. Specifically, the population size can be set to 100, the number of iterations to 50, and a new set of structural parameters can be generated through crossover and mutation operations. Key geometric features, such as the thickness and angle of the support structure, are extracted from the adjusted set of structural parameters. Then, the stress distribution under pressure changes in the inflatable cavity is simulated using finite element analysis. For example, the process of pressure changing from 100 kPa to 120 kPa is simulated to analyze the stress changes at each location. If the maximum stress is less than 70% of the material's yield strength, the adjustment scheme can be judged to have good stability. Finally, the adjustment scheme is verified to meet the compression degree constraint through preset convergence conditions. As one implementation method, the convergence condition can be set as follows: the change in the compression degree value is less than 0.05 in three consecutive iterations. When this condition is met, the final optimized installation structure parameters can be obtained. This method can effectively balance structural stability and space utilization efficiency, improving the installation quality and service life of the LCD screen 7 in the inflatable cavity environment.

[0070] The step of determining the stability of power distribution based on the adjusted installation structure stress data and generating an optimal power distribution model includes:

[0071] The pressure change trend and stress data inside the inflation chamber are analyzed to determine the fluctuation conditions and threshold range. Historical data are classified and trained using a support vector machine, and hyperparameters are adjusted to generate a classification model. Logistic regression is used to analyze historical power allocation data to generate the optimal power allocation model and output the optimization results.

[0072] By acquiring optimized installation location stress data and combining it with the operating status of air pump 3, the internal pressure change trend is analyzed to determine whether it causes fluctuations in power distribution stability, resulting in a preliminary impact assessment. The correlation features between internal pressure and power distribution are extracted from the preliminary impact assessment results. Data analysis techniques are used to process the stress data and historical data to determine the specific conditions and threshold ranges for fluctuations. For the determined fluctuation conditions and threshold ranges, power distribution records related to internal pressure are screened from historical data to obtain a training dataset. The training dataset is trained using a support vector machine (SVM) for classification, and hyperparameters are adjusted and the separating hyperplane is optimized to obtain a classification model capable of distinguishing between stable and unstable states. Based on the classification model, real-time stress data and internal pressure input are analyzed to determine whether the current power distribution is in a stable state, outputting a stability classification result. The classification result is combined with power distribution requirements, and a logistic regression algorithm is used to perform regression analysis on historical power distribution data to obtain the initial parameters of the optimal power distribution model. By inputting real-time internal pressure and stress data into the optimal power distribution model, the power distribution scheme is adjusted, and the final optimized power distribution result is output.

[0073] Specifically, combining the optimized installation location stress data with the operating status of the air pump 3 allows for in-depth analysis of internal pressure change trends. This analysis helps determine whether pressure fluctuations affect power distribution stability, thus obtaining preliminary impact assessment results. As one implementation method, multiple pressure sensors can be installed at different locations inside the inflation chamber to monitor pressure changes in real time. For example, installing one pressure sensor at the top, bottom, and side of the chamber, with a data acquisition frequency of 10 times per second, can accurately capture pressure fluctuations. Specifically, time series analysis methods can be used to process the pressure data, such as using moving averages to smooth short-term fluctuations and then using an autoregressive model to predict pressure trends. This approach can identify abnormal pressure change patterns, such as sudden pressure drops or sustained pressure increases. Furthermore, correlation analysis between pressure data and power distribution data can reveal potential correlations. For example, when the pressure suddenly increases by 5%, a 2% increase in power supply fluctuations is observed. This correlation feature can be extracted using data mining techniques such as association rule learning algorithms. It should be noted that this analysis must consider not only immediate effects but also lag effects, as pressure changes may only affect power supply stability after a certain period. Support Vector Machines (SVMs) are an effective classification algorithm used to distinguish between stable and unstable power distribution states. When training an SVM model, a radial basis function (RBF) can be chosen as the kernel function, and the hyperparameters C and γ can be optimized using cross-validation. For example, C can be set to a range of [0.1, 1, 10, 100], and γ to a range of [0.01, 0.1, 1, 10], finding the optimal parameter combination through grid search. The optimal power distribution model can be constructed using logistic regression. Specifically, internal pressure, force data, etc., can be used as independent variables, and power supply stability as the dependent variable (0 represents unstable, 1 represents stable). The model parameters are solved using maximum likelihood estimation, resulting in a model that can predict the probability of power supply stability. The advantage of this method is that it can quantify the influence of different factors on power supply stability. When adjusting the power distribution scheme in real time, the model prediction results and preset rules can be combined. For example, when the model predicts a power instability probability exceeding 80%, a power reallocation mechanism is triggered. The specific reallocation strategy can be based on priority, allocating more power to critical equipment while reducing power supply to non-critical equipment. This dynamic adjustment method can improve the overall stability and reliability of the system. It is important to emphasize that this data-driven power distribution optimization method has strong adaptability and scalability. With data accumulation and model iteration, the system can continuously learn and improve, adapting to different working environments and equipment configurations. This not only improves power supply efficiency but also extends equipment lifespan and reduces system failures caused by unstable power supply.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A doll with an intelligent simulated eye, characterized by: The device includes a support body; the support body is equipped with simulated eyes and an air pump; the simulated eyes are LCD screens; the air pump is electrically connected to the LCD screen; the support body has an inflation cavity; the output end of the air pump is connected to the inflation cavity.

2. A doll with an intelligent simulated eye according to claim 1, characterized in that: A PVC sheet is provided between the LCD screen and the supporting body.

3. A doll with an intelligent simulated eye according to claim 2, characterized in that: The PVC sheet extends to provide a connecting seat; the connecting seat is detachably connected to a waterproof back cover; a receiving cavity is formed between the connecting seat and the waterproof back cover; the LCD screen is disposed in the receiving cavity; the PVC sheet has a display channel that communicates with the receiving cavity.

4. A doll with an intelligent simulated eye according to claim 3, characterized in that: A waterproof silicone sheet is provided between the LCD screen and the connector; The PVC sheet is connected to the support body by glue and / or sewing.

5. The doll with an intelligent simulated eye according to claim 1, wherein: The doll with intelligent simulated eyes also includes a rotating disk; the support body is located on the rotating disk.

6. The doll with an intelligent simulated eye according to claim 1, wherein: Includes the following steps: S101. Obtain real-time data of the air pump during operation to obtain dynamic correlation data between load fluctuation and power supply distribution; S102. Determine the output status of the air pump based on the dynamic correlation data, and adjust the voltage output range of the air pump. S103. Determine the power supply stability of the LCD screen based on the adjusted voltage output range, and use a voltage regulator module to compensate and adjust the power supply line. S104. Adjust the mounting structure of the LCD screen according to the pressure changes inside the inflation chamber; S105. Determine the stability of power distribution based on the stress data of the adjusted installation structure, and generate the optimal power distribution model.

7. A doll with an intelligent simulated eye according to claim 6, characterized in that: The process of acquiring real-time data during the operation of the air pump to obtain dynamic correlation data between load fluctuations and power supply distribution includes: The current, air pressure, and LCD screen power supply voltage values ​​of the air pump during operation are acquired by sensors. The digital data is processed by time series analysis to determine the frequency components and dynamic correlation strength of load fluctuations. The influence coefficient of voltage stability on power distribution is determined by regression analysis, and the classification model results are generated by support vector machine algorithm.

8. A doll with an intelligent simulated eye according to claim 6, characterized in that: The step of determining the output status of the air pump based on the dynamic correlation data and adjusting the voltage output range of the air pump includes: The output value is extracted from the dynamic association dataset and compared with the preset threshold. If it exceeds the preset range, the target voltage output range value is obtained, the voltage parameters of the air pump drive module are adjusted, the control parameters are optimized using the random forest algorithm, and stable output state data is generated.

9. The doll with an intelligent simulated eye according to claim 6, wherein: The process of determining the power supply stability of the LCD screen based on the adjusted voltage output range, and using a voltage regulator module to compensate and adjust the power supply line, includes: The adjusted voltage output range is compared with the power supply threshold of the LCD screen to extract the characteristics of the unstable state. The adjustment parameters of the voltage regulator module are obtained from the preset module. The deviation characteristics are optimized by the support vector machine algorithm, the control logic of the voltage regulator module is updated, and the stable voltage output data is generated.

10. The doll with an intelligent simulated eye according to claim 6, wherein: The mounting structure for adjusting the LCD screen according to changes in internal pressure of the inflation chamber includes: Real-time pressure data inside the inflation chamber is obtained, and the time series data of pressure change is processed by a convolutional neural network to determine the spatial compression quantification index. The deformation is predicted by a regression algorithm, and the stress distribution is simulated by finite element analysis to verify the stability of the adjustment scheme and obtain the optimized installation structure parameters. The step of determining the stability of power distribution based on the adjusted installation structure stress data and generating an optimal power distribution model includes: The pressure change trend and stress data inside the inflation chamber are analyzed to determine the fluctuation conditions and threshold range. Historical data are classified and trained using a support vector machine, and hyperparameters are adjusted to generate a classification model. Logistic regression is used to analyze historical power allocation data to generate the optimal power allocation model and output the optimization results.

Images