Hug robot for adjuvant therapy of depression

Abstract

An embracing robot for adjuvant therapy of depression belongs to the crossing field of medical health and man-machine interaction and comprises a sensor module, a driving module, a control module, a feedback module, a touch module and a software module. Through the sponge-cotton-sponge interlayer design, the comfortable touch feeling is provided, the warmth and softness of human hugging are simulated, and the emotional satisfaction of a user is enhanced. By combining a sensor and an algorithm, the robot can monitor the embracing strength and duration of the user in real time, the embracing strength and mode are intelligently adjusted, and personalized emotion support is provided. The embracing experience every time is not monotonous due to the randomly changing embracing postures and diversified action modes, the sense of reality and the sense of naturalness are increased, and the mechanical experience is avoided. Through pacifying actions, anxiety and pressure of a user are effectively relieved, treatment of depression is assisted, and emotional comfort is provided.

Description

Technical Field

[0001] The present invention belongs to the intersection of medical health and human-computer interaction, and in particular relates to a hugging robot for auxiliary treatment of depression. Background Art

[0002] Traditional treatments for depression mainly include medication, psychotherapy, and electroconvulsive therapy. Medication usually uses antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), but these drugs may cause side effects such as nausea, headache, and insomnia, and need to be taken for a long time, which affects the patient's life. Psychotherapy, including cognitive behavioral therapy (CBT) and psychoanalysis, can help improve the patient's mental state, but requires the guidance of a professional psychologist, and the course of treatment is long and the cost is high. Although electroconvulsive therapy (ECT) is effective for some patients with refractory depression, it has certain risks and side effects, such as memory loss and impaired consciousness.

[0003] Research shows that comfort dogs can help people with depression in the following ways: Comfort dogs provide unconditional companionship, helping patients reduce loneliness and social isolation. Interacting with animals can reduce the body's stress levels and release endorphins and oxytocin, which can reduce anxiety and depressive symptoms. Patients who own dogs tend to interact with others more frequently and increase their social support network. Some clinical studies have shown the effectiveness of comfort dogs in treating mental health problems such as depression, anxiety, and post-traumatic stress disorder (PTSD). For example, combining psychotherapy with the companionship of comfort dogs can improve the effectiveness of treatment, and patients show better emotional responses when receiving psychological interventions. Comfort dogs can serve as a supplement to drug treatment and reduce patients' dependence on drugs.

[0004] The existing Hug Squeeze Machine was invented by Temple Grandin in 1992 and was originally designed to provide tactile relief to people with autism. Modeled after cow troughs on farms where cows are squeezed to help them calm down, the device features a padded neck opening and a comfortable headrest, and is fully lined with foam rubber. The Hug Squeeze Machine provides tactile stimulation through a manually controlled squeeze, helping the user tolerate specific touch, which for people with autism can help them accept contact with others. The development of this machine has led to the widespread use of this machine, which simulates the squeeze of a hug, in the treatment of children with autism.

[0005] Launched by SoftBank Robotics in 2014, Pepper is an artificial intelligence robot designed for social interaction. Pepper's exterior uses lightweight and soft materials such as foam rubber and silicone rubber to simulate the touch of human skin, and is equipped with tactile sensors and distance sensors that can recognize the user's body language and position to achieve more natural and safe interaction. Its body design also allows it to make some physical contact, including handshakes and hugs. Pepper has been put into use in many countries and regions, including Japan, China, Europe and the United States. It is used in shopping malls, banks, hospitals, schools and other public places to provide information, entertainment and social interaction. Pepper is 1.21 meters tall, which is smaller than the height of an adult, which may limit its ability to provide comfort when hugging, especially for users who need a taller hugging object. And Pepper's main function is social interaction and providing information services. Hugging is only one of its many ways of interaction, so it may not provide the deep interaction and emotional support that a dedicated hugging robot can provide when hugging. Although Pepper's body is made of soft materials, it is not designed specifically for hugging and cannot simulate the comfort of a real hug. It lacks the body temperature, softness and pressure distribution of a real human hug, making it difficult to simulate a real interpersonal hug. Pepper's hug also lacks a personalized experience and cannot be adjusted according to the user's hug needs.

[0006] HuggieBot 3.0, developed by Alexis E. Block and his team, is a robot designed specifically for hugging, with a height and shape designed to mimic the human body. The robot uses soft materials, such as foam and flexible fabric, to cover the robot's exterior to simulate the soft touch of a human hug. It is equipped with advanced sensors and algorithms that can adjust its hugging method according to the user's movements and strength, thereby providing a personalized interactive experience. The design of this robot focuses on providing emotional comfort and is suitable for people who need additional emotional support, such as lonely elderly people, patients, or individuals who need comfort. In addition, as a research robot, HuggieBot 3.0 has not yet been widely used in commercial or home environments. When using HuggieBot, users can only hug while standing, which makes the hug instantaneous. Studies have shown that the duration of the hug will affect the effect. Under the premise that the hug is not too long, the positive effects of the hug may increase over time, and the longer the hug, the stronger the emotional connection between the two parties may feel. However, HuggieBot is not designed to provide a lasting hug experience and emotional comfort. The head of HuggieBot is an electronic screen, which is less human-like and stiff, making it difficult to give people a sense of intimacy. HuggieBot's hugging movements are very slow and the hugging trajectory is rigid, which increases the stiffness of the machine and makes it difficult for users to give users a human-like, simulated, real hugging experience.

[0007] The present invention aims to provide a safe and side-effect-free auxiliary treatment for depression, so as to help patients adjust their mental state during treatment, gradually reduce their dependence on traditional drug treatment, and avoid its side effects. The hugging robot of the present invention can make real-time adjustments according to the user's specific needs for hugging, and enhance the effect and satisfaction of treatment by providing personalized treatment plans. At the same time, the present invention is committed to reducing the discomfort that may occur during treatment and improving the patient's comfort and treatment experience. Summary of the invention

[0008] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0009] In view of the above or existing problems of hugging robots used for auxiliary treatment of depression, the present invention is proposed.

[0010] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0011] The embodiment of the present invention provides a hugging robot for auxiliary treatment of depression, including: a sensor module, which is used to monitor the hugging strength of the user in real time, provide feedback to adjust the hugging strength of the robot, detect the angle and position of the robot relative to the user, and trigger a soothing action when the hugging time of the user exceeds a set threshold; a driving module, which is used to drive a servo to realize the robot's patting and stroking actions, execute the action mode according to a set time period, realize the repetition of patting and stroking, adjust the posture and angle of the robot to adapt to the user's hug, and adjust the action strength to match the user's needs; a control module, which is used to process sensor module data, execute a control algorithm, and optimize the hugging strength and action smoothness through a PID controller; a feedback module, which is used to monitor the hugging strength of the user in real time, judge its changes, and adjust the hugging strength and mode of the robot according to real-time feedback to ensure that it matches the user's needs; a touch module, which is used to adopt a sponge-cotton-sponge sandwich design to provide a comfortable touch and simulate the feeling of a human hug; a software module, which is used to define and schedule the robot's action cycle, randomly change the hugging posture, and enhance the sense of reality.

[0012] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the real-time monitoring of the strength of the user's hug and providing feedback to adjust the robot's hug strength include: setting the sensor data acquisition frequency, converting the output signal of the pressure sensor into a digital signal, using a filtering algorithm to eliminate sensor noise, analyzing the collected pressure data through a microcontroller, calculating the current hugging strength, and comparing it with a preset target strength. The execution cycle of the above program is 100ms.

[0013] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, wherein: the angle and position of the robot relative to the user are detected, and a soothing action is triggered when the user's hugging time exceeds a set threshold, including: using an IMU sensor composed of an accelerometer and a gyroscope to measure the angle and position changes of the robot, regularly collecting data from the IMU sensor, converting the collected angle and position data into a digital signal suitable for processing, and calculating the relative angle and position of the robot relative to the user based on the sensor data; when the user starts to hug, starting a timer to record the duration of the hug in real time, setting a hugging time threshold, and when the user's hugging time exceeds this threshold, triggering subsequent actions, presetting a mode of the soothing action, including specific steps and timing of patting and rubbing the back, once it is detected that the user's hugging time exceeds the set threshold, the control module immediately triggers the execution program of the soothing action, and controls the driving module to implement the specified action.

[0014] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, wherein: the action mode is executed according to the set time period to achieve repetition of patting and stroking, including: setting the duration of the cycle to 20s, ensuring that the duration and interval time of the patting and stroking actions are clear in each cycle, writing a control program in a microcontroller, automatically executing according to the set time cycle and action sequence, and designing action scheduling logic to trigger the execution of patting and stroking at the beginning of each cycle; sending specific drive instructions to the servo to control the movement of the wrist and elbow, the patting action is specifically to set the rotation angle and speed of the servo to achieve four quick pats, and the stroking action is specifically to adjust the rotation range and speed of the servo to perform two up and down strokings.

[0015] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the method of adjusting the posture and angle of the robot to adapt to the hug of the user includes:

[0016] Use IMU sensors to monitor the robot's posture in real time, including roll, pitch and yaw angles, and collect IMU sensor data regularly; use pressure sensors to monitor the user's hug strength to determine whether the user is in a hug state, and set the target posture that the robot needs to achieve based on the user's relative position and posture to ensure that the contact area between the robot and the user is maximized;

[0017] The PID control algorithm is used to adjust the robot's posture in real time to ensure that it is coordinated with the user's hugging action. The PID control algorithm expression is as follows:

[0018]

[0019] Among them, u(t) is the controller output, e(t) is the error at the current moment, K p is the proportional gain, K i is the integral gain, K d is the differential gain;

[0020] Proportional control P means calculating the output based on the error between the current posture and the target posture, which is used to quickly adjust the robot posture; integral control I means accumulating past errors to eliminate long-term steady-state errors and ensure that the robot can maintain the target posture for a long time; differential control D means predicting the changing trend of posture error to prevent overshoot and oscillation during posture adjustment and ensure smooth movement.

[0021] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, wherein: the intensity of the action is adjusted to match the needs of the user, including: based on the PID control algorithm, dynamically adjusting the PID parameters according to the data fed back by the user: if the user's pressure feedback increases, then increase K pIf the user's pressure feedback decreases, then K p Use historical data to calculate the average value and dynamically adjust K i and K d ,To adapt to different user states, it continuously monitors the user's hugging strength and updates the controller input in each control cycle. According to the output of the adaptive PID controller, the robot's action intensity is adjusted in real time.

[0022] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the processing of sensor module data, execution of control algorithm, and optimization of hugging strength and action smoothness through PID controller include:

[0023] Use a pressure sensor to monitor the user's hug strength on the robot in real time, collect pressure values ​​regularly every 100ms, apply a Kalman filter to smooth the sensor data, and calculate the current error e(t), that is, the difference between the target strength and the actual strength:

[0024] e(t) = target force - actual force

[0025] Based on the PID control algorithm, the controller output u(t) is converted into specific instructions for the servo, and the intensity of the slapping and rubbing is adjusted. The output value directly affects the angle and speed of the servo to achieve the required hugging intensity; the interpolation method is used to smoothly transition between actions to avoid sudden movement changes, and the smooth movement of the servo is achieved by setting the movement time and target position;

[0026] Within each control cycle of 20 seconds, the target force is updated in real time based on user feedback, and the control output is recalculated to ensure the consistency and smoothness of the action.

[0027] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the real-time monitoring of the hugging strength of the user, judging its changes, and adjusting the hugging strength and method of the robot according to the real-time feedback to ensure matching with the needs of the user include:

[0028] Use the pressure sensor to continuously collect the user's hug force data, and apply the moving average filter to smooth the collected pressure data to reduce instantaneous fluctuations. The expression is as follows:

[0029]

[0030] Among them, P smoothed (t) is the smoothed pressure value at the current time t, P(t) is the original pressure value at the current time t, N is the window size or the number of samples, and i is the index variable;

[0031] Define fuzzy rules based on the changes in user pressure, set pressure ranges to distinguish "light hug", "moderate hug" and "strong hug". When the user's hug pressure value is in the range of 0-10 units, the robot adopts "light hug", when the user's hug pressure value is in the range of 10-20 units, the robot adopts "moderate hug", when the user's hug pressure value is in the range of 20-30 units, the robot adopts "strong hug";

[0032] Use fuzzy sets to map user inputs to control outputs, with input pressure values ​​such as low, medium, and high, and output adjustment strengths such as decrease, hold, and increase;

[0033] The Mamdani fuzzy inference system is used to calculate the output in combination with the membership function:

[0034] u(t)=Defuzzification(fuz2y_rules)

[0035] Define a state space to represent different hug intensities and methods. The state includes current pressure, user preferences, and historical hug feedback. Give corresponding "rewards" or "penalties" based on user satisfaction, such as user active feedback and pressure duration:

[0036]

[0037] Among them, "reward" means that when the user is satisfied with the robot's hugging experience, or gives a positive evaluation in the user's active feedback, the system regards it as a "reward". When the user satisfaction is high, that is, the sensor detects that the user's hugging strength of the robot continues to increase or the user continues to hug the robot for more than the set time threshold, the robot increases the priority of this behavior; "penalty" means that when the user's feedback is negative, that is, the sensor detects that the user's hugging strength of the robot continues to decrease or the user continues to hug the robot for less than the set time threshold, the system will regard it as a "penalty". When the user satisfaction is low, the robot reduces the frequency of execution of this behavior;

[0038] Use the Q-learning algorithm to update the state-action value function:

[0039]

[0040] Where Q(s,a) is the state-action value function of taking action a in the current state s, s is the current state, a is the currently selected action, α is the learning rate, R is the immediate reward obtained after taking action a in the current state s, γ is the discount factor, s′ is the new state transferred to after executing action a, and a′ is the action that can be taken in the new state s′.

[0041] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the sponge-cotton-sponge sandwich design is adopted to provide a comfortable touch and simulate the feeling of a human hug, including: through the combination of sponge and cotton, the robot maintains temperature characteristics similar to those of the human body to a certain extent, further enhancing the realism of the hug; the combination of sponge and cotton fixes the sandwich material by high-strength stitching or bonding to improve durability during repeated use and reduce the risk of separation or damage; based on the sandwich design, a sensor is integrated to automatically adjust the compression degree of the sponge according to the user's hug strength to provide a personalized hug experience: the sensor monitors in real time The strength of the user's hug on the robot records the pressure applied to the sponge. The collected pressure data is smoothed through a filtering algorithm to obtain a stable pressure value. The control system analyzes the difference between the current pressure value and the preset target pressure value, and calculates the degree of compression that needs to be adjusted. For example, when the pressure applied by the user is less than the target value, the system determines to increase the compression of the sponge; based on the calculation results, the control system sends an instruction to the drive module to adjust the degree of compression of the sponge; the adjustment of the compression degree of the sponge is achieved in the following ways: using an electric servo to control the compression plate or other mechanical structure of the sponge, and realizing compression adjustment according to the instruction; adjusting the compression range of the sponge by a servo motor to adapt to the strength of the user's hug.

[0042] As a preferred solution of the hugging robot for auxiliary treatment of depression described in the present invention, the definition and scheduling of the robot's action cycle and random changes in the hugging posture include: in each action cycle, the robot randomly selects the arm servo angle change within a certain range to achieve the randomness of the hugging posture, and the robot shows a different hugging gesture each time it extends its hand. By using a random number generator, new arm servo angle parameters are generated according to the set angle range to ensure that each hugging gesture has changes; a timer or state machine is used to schedule the action cycle. At the end of each cycle, the robot changes according to the pressure applied by the user through hugging, and the sensor monitors and records the pressure value in real time. The system compares the pressure value with the target pressure value to determine whether to execute the next cycle to determine the user's satisfaction or the length of time the user continues to hug the robot.

[0043] The beneficial effects of the present invention are as follows: the present invention provides a comfortable touch through the interlayer design of sponge-cotton-sponge, simulates the warmth and softness of a human hug, and enhances the user's emotional satisfaction. Combined with sensors and algorithms, the robot can monitor the user's hug strength and duration in real time, intelligently adjust the hug strength and method, and provide personalized emotional support. The randomly changing hug postures and diversified action patterns make each hug experience no longer monotonous, increase the sense of reality and naturalness, and avoid a mechanized experience. Through soothing movements, the user's anxiety and stress are effectively relieved, the treatment of depression is assisted, and emotional comfort is provided. Through the positive and negative feedback mechanism, it is ensured that the robot safely adjusts the hug strength according to the user's feedback to avoid discomfort or harm to the user. The robot can dynamically adjust according to the individual differences and emotional changes of the user, adapt to different usage scenarios and user needs, and improve the flexibility of application. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0045] Figure 1 A schematic diagram of the structure of a hugging robot for auxiliary treatment of depression provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0047] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0048] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0049] Example 1

[0050] Reference Figure 1, is an embodiment of the present invention, which provides a hugging robot for auxiliary treatment of depression, including:

[0051] S1: The sensor module is used to monitor the strength of the user's hug in real time, provide feedback to adjust the robot's hug strength, detect the angle and position of the robot relative to the user, and trigger a soothing action when the user's hug time exceeds a set threshold.

[0052] Preferably, the sensor data acquisition frequency is set, the output signal of the pressure sensor is converted into a digital signal, a filtering algorithm is used to eliminate sensor noise, the collected pressure data is analyzed by a microcontroller, the current hugging force is calculated, and compared with a preset target force.

[0053] Preferably, an IMU sensor composed of an accelerometer and a gyroscope is used to measure the angle and position changes of the robot, the data of the IMU sensor is collected regularly, the collected angle and position data are converted into a digital signal suitable for processing, and the relative angle and position of the robot relative to the user are calculated based on the sensor data;

[0054] When the user starts to hug, the timer is started to record the duration of the hug in real time and set a hug time threshold. When the user's hug time exceeds this threshold, the subsequent action is triggered and the mode of the soothing action is preset, including the specific steps and timing of patting and stroking the back. Once it is detected that the user's hug time exceeds the set threshold, the control module immediately triggers the execution program of the soothing action and controls the driving module to implement the specified action.

[0055] Furthermore, the sensor data acquisition frequency is set to 10Hz, and the sampling frequency should be adjusted according to the actual usage scenario to ensure real-time response to force changes; the analog signal output by the sensor is converted into a digital signal through ADC, and the resolution is selected to be 12-bit or 16-bit ADC, providing high-precision pressure detection, which can effectively reflect subtle force changes, and map the voltage range of the sensor output (such as 0-5V) to the corresponding digital value (such as 0-4095) for subsequent data processing;

[0056] The microcontroller receives the filtered pressure data in real time and calculates the current average pressure (i.e. the user's hugging strength). The actual hugging strength is obtained through weighted mean. A preset target hugging strength is stored in the system.

[0057] Through the control algorithm in the microcontroller, the current hugging force Pcurrent is compared with the target hugging force Ptarget. The control logic is:

[0058] If Pcurrent > Ptarget, the robot appropriately reduces the pressure, for example, by loosening the servo or reducing the motor output torque to decrease the hugging intensity.

[0059] If Pcurrent < Ptarget, the robot appropriately increases the pressure, for example, by enhancing the servo strength or increasing the motor output torque to make the hug tighter.

[0060] Furthermore, the data acquisition frequency is set to 50Hz, and data is read from the IMU sensor every 20ms, including three-axis acceleration and three-axis angular velocity.

[0061] The collected analog signal is converted into a digital signal through ADC, and through a filtering algorithm (to eliminate noise and obtain smooth angle and position data; based on the data of the accelerometer, calculate the tilt angle and displacement information of the robot; combined with the angular velocity data of the gyroscope, use a fusion algorithm to accurately estimate the posture of the robot in three-dimensional space, thereby obtaining the relative angle and position of the robot with respect to the user; when the user starts to hug, by detecting the signal of the pressure sensor, determine that the user's arms surround the robot and start the timer to record the duration of the hug.

[0062] Set 20s as the threshold for the hug duration. If the user's hug duration exceeds this threshold, it is considered that the user needs more emotional support, and the robot will automatically enter the soothing mode.

[0063] Definition of the soothing action mode: The action cycle is defined as 20s, including two main soothing actions: patting and rubbing the back.

[0064] Steps of the action cycle:

[0065] Patting action: Use the servo at the wrist part to quickly and gently pat the user's back 4 times;

[0066] Rubbing action: Rotate the shaft on the servo to drive the arm movement to achieve 2 up-and-down rubs on the user's back;

[0067] Repeating process: Pat 4 times - Rub 2 times to complete a full soothing action cycle;

[0068] Continuously monitor the timer through the microcontroller. Once the user's hug duration exceeds the set 20s threshold, the microcontroller will immediately trigger the soothing action execution program; when the trigger condition is met, the control module calls the preset action program to control the robot to perform periodic soothing actions; through the servo control unit, the driving module transmits the instructions to the servos at the robot's wrist and elbow to execute specific patting and rubbing actions; the robot repeats the above soothing actions every 20s until the user stops hugging.

[0069] S2: The drive module is used to drive the servo to realize the robot's patting and stroking actions, execute the action mode according to the set time period, realize the repetition of patting and stroking, adjust the robot's posture and angle to adapt to the user's hug, and adjust the intensity of the action to match the user's needs.

[0070] Preferably, the duration of the cycle is set to 20s. In each cycle, the duration and interval of the tapping and rubbing actions are ensured to be clear. A control program is written in the microcontroller to automatically execute according to the set time cycle and action sequence, and the action scheduling logic is designed to trigger the execution of tapping and rubbing at the beginning of each cycle.

[0071] Send specific drive instructions to the servo to control the movement of the wrist and elbow. The flapping action is to set the rotation angle and speed of the servo, and implement four quick flaps. The rubbing action is to adjust the rotation range and speed of the servo, and perform two up and down rubbings.

[0072] Preferably, an IMU sensor is used to monitor the robot's posture in real time, including roll, pitch and yaw angles, and IMU sensor data is collected regularly; a pressure sensor is used to monitor the user's hugging strength to determine whether the user is in a hugging state, and a target posture that the robot needs to achieve is set according to the user's relative position and posture to ensure that the contact area between the robot and the user is maximized;

[0073] The PID control algorithm is used to adjust the robot's posture in real time to ensure that it is coordinated with the user's hugging action. The PID control algorithm expression is as follows:

[0074]

[0075] Among them, u(t) is the controller output, e(t) is the error at the current moment, K p is the proportional gain, K i is the integral gain, K d is the differential gain;

[0076] Proportional control P means calculating the output based on the error between the current posture and the target posture, which is used to quickly adjust the robot posture; integral control I means accumulating past errors to eliminate long-term steady-state errors and ensure that the robot can maintain the target posture for a long time; differential control D means predicting the changing trend of posture error to prevent overshoot and oscillation during posture adjustment and ensure smooth movement.

[0077] Preferably, based on the PID control algorithm, the PID parameters are dynamically adjusted according to the data fed back by the user:

[0078] If the user's pressure feedback increases, increase K p The value of

[0079] If the user's pressure feedback decreases, reduce K p The value of

[0080] Use historical data to calculate the average value and dynamically adjust K i and K d ,To adapt to different user states, it continuously monitors the user's hugging strength and updates the controller input in each control cycle. According to the output of the adaptive PID controller, the robot's action intensity is adjusted in real time.

[0081] Furthermore, the cycle of the entire soothing action is set to 20 seconds, and a complete sequence of patting and stroking actions is completed in each cycle.

[0082] Action sequence and interval time:

[0083] Patting action: 4 quick taps, each tap lasts 1 second, and the interval between taps is 0.5 seconds;

[0084] Rubbing action: rub up and down 2 times, each rubbing lasts 4 seconds, and the interval between rubbing is 1 second;

[0085] The microcontroller is responsible for the scheduling and execution of the entire soothing action, which is achieved through timed tasks and precise control. When the robot starts, the servo control module is initialized, the initial position and angle of the servo are set, and the robot is in standby mode. Set a periodic timer to perform the soothing action regularly, triggering it every 20 seconds. Whenever the periodic timer is triggered, the scheduler executes the following steps: Patting action: Complete 4 quick pats by sending the servo rotation command.

[0086] Interval: Wait 0.5s between each tap;

[0087] Stroking action: complete up and down stroking 2 times, each with a 1s interval;

[0088] Waiting cycle ends: After completing the action sequence, wait for the remaining time to reach 20s.

[0089] S3: The control module is used to process the sensor module data, execute the control algorithm, and optimize the hugging force and movement smoothness through the PID controller.

[0090] Preferably, a pressure sensor is used to monitor the hugging force of the user on the robot in real time, and the pressure value is collected regularly every 100ms. The sensor data is smoothed by applying a Kalman filter to calculate the current error e(t), that is, the difference between the target force and the actual force:

[0091] e(t) = target force - actual force

[0092] Based on the PID control algorithm, the controller output u(t) is converted into specific instructions for the servo, and the intensity of the slapping and rubbing is adjusted. The output value directly affects the angle and speed of the servo to achieve the required hugging intensity; the interpolation method is used to smoothly transition between actions to avoid sudden movement changes, and the smooth movement of the servo is achieved by setting the movement time and target position;

[0093] Within each control cycle of 20 seconds, the target force is updated in real time based on user feedback, and the control output is recalculated to ensure the consistency and smoothness of the action.

[0094] Furthermore, the pressure data collected by the sensor is smoothed using a Kalman filter to eliminate instantaneous fluctuations. The update process of the Kalman filter is as follows: the current state is predicted based on the state at the previous moment, and the state estimate is updated based on the current measurement value;

[0095] The formula is as follows:

[0096]

[0097] K=P-1H(HPA-1H+R)

[0098]

[0099] P:=(IK,H)Pk-1

[0100] Convert the PID control output u(t) into specific instructions for the servo, including adjusting the intensity of slapping and rubbing. The output value will affect the angle and speed of the servo to achieve the desired hug intensity;

[0101] Use interpolation methods to make smooth transitions between actions and avoid abrupt changes in motion.

[0102] Set the movement time tmove and the target position θtarget, and calculate the smooth movement path of the servo within a given time:

[0103]

[0104] Every 20 seconds is a control cycle. In each control cycle, the system will update the target force in real time based on user feedback and recalculate the control output.

[0105] S4: The feedback module is used to monitor the user's hugging strength in real time, determine its changes, and adjust the robot's hugging strength and method based on real-time feedback to ensure that it matches the user's needs.

[0106] Specifically, a pressure sensor is used to continuously collect the user's hugging force data, and a moving average filter is applied to smooth the collected pressure data to reduce instantaneous fluctuations. The expression is as follows:

[0107]

[0108] Among them, P smoothed (t) is the smoothed pressure value at the current time t, P(t) is the original pressure value at the current time t, N is the window size or the number of samples, and i is the index variable;

[0109] Define fuzzy rules such as "light hug", "moderate hug" and "strong hug" according to the changes in user pressure, and use fuzzy sets to map user input to control output, with input pressure values ​​such as low, medium and high, and output adjustment strength such as decrease, maintain and increase;

[0110] The Mamdani fuzzy inference system is used to calculate the output in combination with the membership function:

[0111] u(t)=Defuzzification(fuzzy_rules)

[0112] Define a state space to represent different hug intensities and methods. The state includes current pressure, user preferences, and historical hug feedback. Give corresponding rewards or penalties based on user satisfaction, such as user active feedback and pressure duration:

[0113]

[0114] Among them, "reward" means that when the user is satisfied with the robot's hugging experience, or gives a positive evaluation in the user's active feedback, the system regards it as a "reward". When the user satisfaction is high, that is, the sensor detects that the user's hugging strength of the robot continues to increase or the user continues to hug the robot for more than the set time threshold, the robot increases the priority of this behavior; "penalty" means that when the user's feedback is negative, that is, the sensor detects that the user's hugging strength of the robot continues to decrease or the user continues to hug the robot for less than the set time threshold, the system will regard it as a "penalty". When the user satisfaction is low, the robot reduces the frequency of execution of this behavior;

[0115] Use the Q-learning algorithm to update the state-action value function:

[0116]

[0117] Where Q(s,a) is the state-action value function of taking action a in the current state s, s is the current state, a is the currently selected action, α is the learning rate, R is the immediate reward obtained after taking action a in the current state s, γ is the discount factor, s′ is the new state transferred to after executing action a, and a′ is the action that can be taken in the new state s′.

[0118] Further, if the pressure is "low", the intensity is adjusted to "increase";

[0119] If the pressure is "Medium", adjust the force to "Hold";

[0120] If the pressure is "high", adjust the force to "reduce";

[0121] Use membership functions to map inputs to fuzzy sets, such as the membership function μlow(P) for low pressure:

[0122]

[0123] According to the input membership, the output membership is obtained through fuzzy reasoning, and finally the centroid method is used to defuzzify and obtain the specific adjustment strength;

[0124] Reward or punish users based on their satisfaction:

[0125] If active user feedback (such as smiling or tapping the robot) is considered positive feedback, increase the reward;

[0126] If the user maintains the same hug strength for more than the set time threshold, a reward is given to encourage the robot to maintain the current posture;

[0127] If the user releases the hug or shows discomfort, a penalty will be imposed and the reward value will be reduced.

[0128] S5: The touch module is used to adopt a sponge-cotton-sponge sandwich design to provide a comfortable touch and simulate the feeling of a human hug.

[0129] Preferably, through the combination of sponge and cotton, the robot maintains temperature characteristics similar to those of the human body to a certain extent, further enhancing the realism of the hug; the sandwich material is fixed by high-strength stitching or bonding to ensure that it will not separate or be damaged during repeated use; based on the sandwich design, an integrated sensor can automatically adjust the compression degree of the sponge according to the user's hug strength, providing a personalized hug experience.

[0130] Furthermore, the present invention adopts a novel filling material design scheme to achieve a more human-like and comfortable simulated hugging touch. Sponge can meet the requirements of softness and comfort, and has a certain resilience, while cotton is soft, and has higher softness and comfort than sponge, but has poor resilience and real feel. In order to simulate the rich touch of complex human tissue, by combining multiple groups of materials, making samples and conducting data statistics on the subjects, the sponge-cotton stacking combination with the highest human-like degree and comfort was selected as the material for the robot. The following are 9 different samples made:

[0131]

[0132] S6: The software module is used to define and schedule the robot's motion cycle, randomly changing the hugging posture to enhance the sense of reality.

[0133] Preferably, in each action cycle, the robot randomly selects posture changes within a certain range, randomly adjusts the robot's tilt angle relative to the user, slightly adjusts the distance between the robot and the user within a certain range, and uses a random number generator to generate new posture parameters according to the set angle and position range; a timer or state machine is used to schedule the action cycle, and each time a cycle ends, the robot decides whether to execute the next cycle based on the current state and user feedback.

[0134] Furthermore, the core code of the present invention is as follows:

[0135] Code snippet 1main.c

[0136]

[0137]

[0138]

[0139]

[0140] Code snippet 2 ControlAlgorithm.c

[0141]

[0142]

[0143]

[0144]

[0145]

[0146] 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 present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A hugging robot for auxiliary treatment of depression, characterized in that: include: The sensor module is used to monitor the strength of the user's hug in real time, provide feedback to adjust the robot's hug strength, detect the angle and position of the robot relative to the user, and trigger a soothing action when the user's hug time exceeds a set threshold; The driving module is used to drive the servo to realize the robot's patting and rubbing actions, execute the action mode according to the set time period, realize the repetition of patting and rubbing, adjust the robot's posture and angle to adapt to the user's hug, and adjust the intensity of the action to match the user's needs; The control module is used to process the sensor module data, execute the control algorithm, and optimize the hugging force and movement smoothness through the PID controller; The feedback module is used to monitor the user's hugging strength in real time, judge its changes, and adjust the robot's hugging strength and method based on real-time feedback to ensure that it matches the user's needs; The touch module is used to provide a comfortable touch using a sponge-cotton-sponge sandwich design to simulate the feeling of a human hug; Software module for defining and scheduling the robot's motion cycles, randomly changing hugging postures, and enhancing realism.

2. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The real-time monitoring of the strength of the user's hug and providing feedback to adjust the robot's hug strength include: Set the sensor data acquisition frequency, convert the output signal of the pressure sensor into a digital signal, use the filtering algorithm to eliminate the sensor noise, analyze the collected pressure data through the microcontroller, calculate the current hugging force, and compare it with the preset target force. The execution cycle of the above program is 100ms.

3. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The detection robot detects the angle and position of the robot relative to the user, and triggers a soothing action when the user's hug time exceeds a set threshold, including: Use an IMU sensor that combines an accelerometer and a gyroscope to measure the angle and position changes of the robot, collect data from the IMU sensor at regular intervals, convert the collected angle and position data into digital signals suitable for processing, and calculate the relative angle and position of the robot relative to the user based on the sensor data; When the user starts to hug, the timer is started to record the duration of the hug in real time and set a hug time threshold. When the user's hug time exceeds this threshold, the subsequent action is triggered and the mode of the soothing action is preset, including the specific steps and timing of patting and stroking the back. Once it is detected that the user's hug time exceeds the set threshold, the control module immediately triggers the execution program of the soothing action and controls the driving module to implement the specified action.

4. The hugging robot for auxiliary treatment of depression according to claim 1, characterized in that: The action mode is executed according to the set time period to achieve the repetition of patting and rubbing, including: Set the duration of the cycle to 20s. In each cycle, ensure that the duration and interval of the tapping and rubbing actions are clear. Write a control program in the microcontroller to automatically execute according to the set time cycle and action sequence. Design the action scheduling logic to trigger the execution of tapping and rubbing at the beginning of each cycle. Send specific drive instructions to the servo to control the movement of the wrist and elbow. The flapping action is to set the rotation angle and speed of the servo, and implement four quick flaps. The rubbing action is to adjust the rotation range and speed of the servo, and perform two up and down rubbings.

5. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The adjusting the posture and angle of the robot to adapt to the hug of the user comprises: Use IMU sensors to monitor the robot's posture in real time, including roll, pitch and yaw angles, and collect IMU sensor data regularly; use pressure sensors to monitor the user's hug strength to determine whether the user is in a hug state, and set the target posture that the robot needs to achieve based on the user's relative position and posture to ensure that the contact area between the robot and the user is maximized; The PID control algorithm is used to adjust the robot's posture in real time to ensure that it is coordinated with the user's hugging action. The PID control algorithm expression is as follows: Among them, u(t) is the controller output, e(t) is the error at the current moment, K p is the proportional gain, K i is the integral gain, K d is the differential gain; Proportional control P means calculating the output based on the error between the current posture and the target posture, which is used to quickly adjust the robot posture; integral control I means accumulating past errors to eliminate long-term steady-state errors and ensure that the robot can maintain the target posture for a long time; differential control D means predicting the changing trend of posture error to prevent overshoot and oscillation during posture adjustment and ensure smooth movement.

6. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The intensity of the action is adjusted to match the needs of the user, including: Based on the PID control algorithm, the PID parameters are dynamically adjusted according to the data fed back by the users: If the user's pressure feedback increases, increase K p The value of If the user's pressure feedback decreases, reduce K p The value of Use historical data to calculate the average value and dynamically adjust K i and K d ,To adapt to different user states, it continuously monitors the user's hugging strength and updates the controller input in each control cycle. According to the output of the adaptive PID controller, the robot's action intensity is adjusted in real time.

7. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The processing of sensor module data, execution of control algorithms, and optimization of hugging strength and motion smoothness through a PID controller include: Use a pressure sensor to monitor the user's hug strength on the robot in real time, collect pressure values ​​regularly every 100ms, apply a Kalman filter to smooth the sensor data, and calculate the current error e(t), that is, the difference between the target strength and the actual strength: e(t) = target force - actual force Based on the PID control algorithm, the controller output u(t) is converted into specific instructions for the servo, and the intensity of the slapping and rubbing is adjusted. The output value directly affects the angle and speed of the servo to achieve the required hugging intensity; the interpolation method is used to smoothly transition between actions to avoid sudden movement changes, and the smooth movement of the servo is achieved by setting the movement time and target position; Within each control cycle of 20 seconds, the target force is updated in real time based on user feedback, and the control output is recalculated to ensure the consistency and smoothness of the action.

8. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The real-time monitoring of the hugging strength of the user, judging its changes, and adjusting the hugging strength and manner of the robot according to the real-time feedback to ensure matching with the needs of the user include: Use the pressure sensor to continuously collect the user's hug force data, and apply the moving average filter to smooth the collected pressure data to reduce instantaneous fluctuations. The expression is as follows: Among them, P smoothed (t) is the smoothed pressure value at the current time t, P(t) is the original pressure value at the current time t, N is the window size or the number of samples, and i is the index variable; Define fuzzy rules based on the changes in user pressure, set pressure ranges to distinguish "mild hug", "moderate hug" and "strong hug". When the user's hug pressure value is within the range of 0-10 units, the robot adopts "mild hug", when the user's hug pressure value is within the range of 10-20 units, the robot adopts "moderate hug", when the user's hug pressure value is within the range of 20-30 units, the robot adopts "strong hug"; Use fuzzy sets to map user inputs to control outputs, with input pressure values ​​such as low, medium, and high, and output adjustment strengths such as decrease, hold, and increase; The Mamdani fuzzy inference system is used to calculate the output in combination with the membership function: u(t)=Defuzzification(fuzzy_rules) Define the state space to represent different hugging intensities and methods. The state includes current pressure, user preferences, and historical hugging feedback. Give corresponding "rewards" or "penalties" based on user satisfaction, such as user active feedback and pressure maintenance time: Among them, "reward" means that when the user is satisfied with the robot's hugging experience, or gives a positive evaluation in the user's active feedback, the system regards it as a "reward". When the user satisfaction is high, that is, the sensor detects that the user's hugging strength of the robot continues to increase or the user continues to hug the robot for more than the set time threshold, the robot increases the priority of this behavior; "penalty" means that when the user's feedback is negative, that is, the sensor detects that the user's hugging strength of the robot continues to decrease or the user continues to hug the robot for less than the set time threshold, the system will regard it as a "penalty". When the user satisfaction is low, the robot reduces the frequency of execution of this behavior; Use the Q-learning algorithm to update the state-action value function: Where Q(s,a) is the state-action value function of taking action a in the current state s, s is the current state, a is the currently selected action, α is the learning rate, R is the immediate reward obtained after taking action a in the current state s, γ is the discount factor, s′ is the new state transferred to after executing action a, and a′ is the action that can be taken in the new state s′.

9. The hugging robot for auxiliary treatment of depression as claimed in claim 1, characterized in that: The sponge-cotton-sponge sandwich design provides a comfortable touch and simulates the feeling of a human hug, including: Through the combination of sponge and cotton, the robot maintains temperature characteristics similar to those of the human body to a certain extent, further enhancing the realism of the hug; the combination of sponge and cotton fixes the sandwich material by high-strength stitching or bonding to improve durability during repeated use and reduce the risk of separation or damage; Based on the sandwich design, the sensor is integrated to automatically adjust the compression of the sponge according to the user's hugging strength, providing a personalized hugging experience: the sensor monitors the user's hugging strength on the robot in real time, records the pressure applied to the sponge, and the collected pressure data is smoothed through a filtering algorithm to obtain a stable pressure value. The control system analyzes the difference between the current pressure value and the preset target pressure value and calculates the compression degree that needs to be adjusted. For example, when the pressure applied by the user is less than the target value, the system determines to increase the compression of the sponge; based on the calculation results, the control system sends instructions to the drive module to adjust the compression degree of the sponge; the compression degree of the sponge is adjusted in the following ways: Use electric servos to control the compression plate or other mechanical structure of the sponge to adjust the compression according to the instructions; The compression range of the sponge is adjusted by a servo motor to adapt to the user's hug strength.

10. The hugging robot for auxiliary treatment of depression according to claim 1, characterized in that: The definition and scheduling of the robot's action cycle and the random change of hugging posture include: In each action cycle, the robot randomly selects arm servo angle changes within a certain range to achieve randomness in the hugging posture. Each time the robot extends its hand, it shows a different hugging gesture. By using a random number generator, new arm servo angle parameters are generated according to the set angle range to ensure that each hugging gesture has changes. A timer or state machine is used to schedule the action cycle. At the end of each cycle, the robot changes according to the pressure applied by the user through hugging. The sensor monitors and records the pressure value in real time. The system compares the pressure value with the target pressure value to determine the user's satisfaction or the length of time the user continues to hug the robot to decide whether to execute the next cycle.

Images