Design method and device for self-thinking robot with human characters

Abstract

The invention relates to the technical field related to robots, in particular to a design method and device for a robot with human characters and capable of achieving self thinking. The method comprises the following steps: acquiring a five-recognition information matrix; wherein the five-recognition information matrix comprises image recognition information, voice recognition information, taste recognition information, smell recognition information and body touch and limb movement recognition information; the five-recognition information matrix is a partial matrix in the state matrix; determining a stimulation matrix based on the change of the state matrix over time and a predetermined character matrix; wherein the character matrix is used for representing characters of the robot; wherein the stimulation matrix is used for representing the change of the state matrix, and stimulation is obtained under the influence of the character matrix; adjusting a state matrix in combination with the stimulation matrix, the character matrix and a preset memory warehouse; the state matrix is used for representing the current state of the robot; based on the adjusted state matrix, behaviors needing to be carried out by the robot are determined, and a behavior matrix is obtained; and adjusting the character matrix in combination with the state matrix and the stimulation matrix.

Description

Technical Field

[0001] This application relates to the technical field of robots, and specifically relates to a design method and device for a robot with human-like personality and self-thinking ability. Background Art

[0002] In today's era of rapid technological development, the field of artificial intelligence has achieved many remarkable achievements, especially the development of robot technology. As robots gradually enter various scenarios such as homes, they have shown great potential in replacing human labor and have become an important trend in future social development. For example, in industrial production, robots can accurately and efficiently complete repetitive labor, improving production efficiency; in home services, they can undertake tasks such as cleaning and nursing, bringing convenience to people's lives.

[0003] Current artificial intelligence mainly focuses on simulating human recognition abilities of the world, especially in the aspects of images and voices. However, this simulation has many deficiencies. On the one hand, the recognition of images and the deeper meaning of images, the generation of voices and the deeper meaning of voices are processed mixed together, which greatly increases the complexity of processing. For example, when recognizing an image containing people and scenes, not only the basic elements of people and scenes need to be recognized, but also the deeper meanings such as emotions and stories that may be contained behind the image need to be understood. Currently, the technology often seems inadequate in dealing with these complex information. On the other hand, in the simulation of human personality and creativity, currently most use convolutional neural networks (CNNs), but this method is relatively rough and basically belongs to a swallow-whole type of simulation, making it difficult to truly achieve accurate simulation of human personality and creativity. For example, when simulating human personality, CNNs may only be able to capture surface behavior patterns and cannot deeply understand the internal mechanisms and complex psychological factors of personality formation.

[0004] The existing technology has obvious limitations in enabling robots to have human-like personalities and thinking abilities, and there is an urgent need for a new method to overcome these problems, so as to achieve better integration of robots with human society. This application is exactly an innovative solution proposed based on such a background. Summary of the Invention

[0005] In view of this, the embodiments of this application are committed to providing a design method and device for a robot with human-like personality and self-thinking ability to analyze and simulate robots more comprehensively and effectively.

[0006] This application provides a design method for a robot with human-like personality and self-thinking ability, including:

[0007] Obtain a five-sense information matrix;

[0008] Among them, the five-sense information matrix includes: image recognition information, sound recognition information, taste recognition information, olfactory recognition information, body touch and limb movement recognition information; the five-sense information matrix is a partial matrix inside the state matrix;

[0009] Based on the change of the state matrix over time and the pre-determined personality matrix, determine the stimulus matrix;

[0010] Among them, the personality matrix is used to characterize the personality of the robot;

[0011] Among them, the stimulus matrix is used to characterize the change of the state matrix and the stimulus obtained under the influence of the personality matrix;

[0012] Combine the stimulus matrix, the personality matrix and the preset memory warehouse to adjust the state matrix;

[0013] The state matrix is used to characterize the current state of the robot and the external environment;

[0014] Based on the adjusted state matrix, determine the actions that the robot needs to perform to obtain the action matrix;

[0015] Combine the state matrix and the stimulus matrix to adjust the personality matrix;

[0016] The memory warehouse is constructed based on the change of the state matrix.

[0017] In some embodiments, it further includes:

[0018] Analyze the change process of the state matrix with relevant relationships to obtain an analysis result;

[0019] The relevant relationship includes a causal relationship;

[0020] Based on the analysis result, update the memory warehouse.

[0021] In some embodiments, updating the memory warehouse includes:

[0022] Associate, strengthen, merge and delete the historical event information in the memory warehouse.

[0023] In some embodiments, it further includes:

[0024] Based on the principle of giving priority to ensuring that the robot generates an action matrix, perform the allocation of calculation examples.

[0025] In some embodiments, the actions include: language, thoughts or actions.

[0026] In some embodiments, it further includes: obtaining a personality description;

[0027] Generate the personality matrix based on the personality description.

[0028] In some embodiments, based on the adjusted state matrix, determine the actions that the robot needs to perform to obtain an action matrix, including:

[0029] If the internal parameters of the adjusted state matrix meet the preset requirements, perform the actions corresponding to the preset requirements;

[0030] Determine the action matrix based on the actions corresponding to the preset requirements.

[0031] In some embodiments, it further includes:

[0032] Automatically update the "self" matrix in the state matrix over time;

[0033] Wherein the "self" matrix is a parameter in the state matrix used to characterize the state of the robot and changes over time.

[0034] The present application provides a robot design device with human-like personality and self-thinking ability, including:

[0035] An acquisition module for acquiring a five-sense information matrix;

[0036] Wherein, the five-sense information matrix includes: image recognition information, sound recognition information, taste recognition information, olfactory recognition information, body touch and limb movement recognition information; the five-sense information matrix is a partial matrix inside the state matrix;

[0037] A first determination module for determining a stimulus matrix based on the change of the state matrix over time and a pre-determined personality matrix;

[0038] Wherein, the personality matrix is used to characterize the personality of the robot;

[0039] Wherein, the stimulus matrix is used to characterize the change of the state matrix and the stimulus obtained under the influence of the personality matrix;

[0040] A combination module for adjusting the state matrix by combining the stimulus matrix, the personality matrix and a preset memory repository;

[0041] The state matrix is used to characterize the current state of the robot and the external environment;

[0042] A second determination module for determining the actions that the robot needs to perform based on the adjusted state matrix to obtain an action matrix;

[0043] An adjustment module for adjusting the personality matrix by combining the state matrix and the stimulus matrix;

[0044] The memory repository is constructed based on the change of the state matrix.

[0045] The present application provides an electronic device, including:

[0046] a processor, and a memory for storing programs executable by the processor;

[0047] The processor is configured to implement the robot design method with a human-like personality and self-thinking as described above by running the programs in the memory.

[0048] A robot design method with a human-like personality and self-thinking provided by the present application obtains a five-sense information matrix; wherein, the five-sense information matrix includes: image recognition information, sound recognition information, taste recognition information, olfactory recognition information, body touch and limb movement recognition information; the five-sense information matrix is a partial matrix inside the state matrix; based on the change of the state matrix over time and a pre-determined personality matrix, a stimulus matrix is determined; wherein, the personality matrix is used to characterize the personality of the robot; wherein, the stimulus matrix is used to characterize the change of the state matrix, the stimulus obtained under the influence of the personality matrix; combining the stimulus matrix, the personality matrix and a preset memory warehouse, the state matrix is adjusted; the state matrix is used to characterize the current state of the robot and the external environment; based on the adjusted state matrix, the behavior that the robot needs to perform is determined to obtain a behavior matrix; combining the state matrix and the stimulus matrix, the personality matrix is adjusted; the memory warehouse is constructed based on the change of the state matrix. With such a setting, the five-sense information matrix is directly obtained instead of processing the information, which simplifies the difficulty of information processing. Combining the physical and social states of the robot itself, a comprehensive state matrix can be formed. This matrix not only reflects the immediate state of the robot, but also can be dynamically adjusted according to environmental changes. Through the preset stimulus matrix and personality matrix, the robot can make personalized responses to external stimuli according to its personality characteristics, which makes the behavior of the robot more in line with its "personality". The behavior of the robot is no longer a preset fixed pattern, but is dynamically adjusted according to the real-time state matrix, making the behavior of the robot more natural and adaptable. The establishment of the memory warehouse enables the robot to store and process past experiences. Through operations such as association, reinforcement, merging and deletion, the robot can learn from experience and improve its behavior and decision-making. The present application provides a more intelligent, personalized and adaptable behavior decision-making mechanism for the robot by simulating human perception, personality and memory, which not only improves the functionality of the robot, but also provides new possibilities for the interaction between the robot and humans. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The above and other objects, features, and advantages of the present application will become more apparent by describing the embodiments of the present application in more detail with reference to the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation to the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0050] Figure 1 It is a schematic flow chart of a robot design method with human-like personality and self-thinking provided by an embodiment of the present application.

[0051] Figure 2 It is an overall operation diagram of thinking provided by an embodiment of the present application.

[0052] Figure 3 It is a personality matrix diagram provided by an embodiment of the present application.

[0053] Figure 4 It is a comprehensive processing process diagram provided by an embodiment of the present application.

[0054] Figure 5 It is a stimulus response diagram provided by an embodiment of the present application.

[0055] Figure 6 It is a schematic diagram of the processing during sleep provided by an embodiment of the present application.

[0056] Figure 7 It is a schematic structural diagram of a robot design device with human-like personality and self-thinking provided by an embodiment of the present application.

[0057] Figure 8 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0058] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0059] Application Overview

[0060] With the gradual development of AI computing based on neural networks towards intelligence, it is possible that robots will enter households and become machines to replace human labor in the future. Just as presented in the TV series Humans, when robots enter households, some social problems may follow. Endowing robots with human personalities seems to be able to solve these social problems.

[0061] The problem with current AI technologies based on neural networks is that the computational workload is huge. For ChatGPT or similar solutions, even when tens of thousands of GPUs work simultaneously, they can only simulate less than 1% of the human brain, and the power consumption is on a celestial scale. Compared to humans who only need the energy from three meals a day to supply the energy consumption of the brain and body for a day, the current method of stacking a large number of GPU chips for the neural network can be said to be quite primitive. With the current method, it is almost unimaginable to enable a robot to have human-like thinking. Moreover, current artificial intelligence mainly simulates humans based on the human's recognition ability of the world. When it comes to simulating human personality and creativity, it is almost all done using CNNs, which is basically just swallowing everything whole.

[0062] On the other hand, current artificial intelligence's simulation of human thinking focuses on images and speech. The recognition of images and the deeper meaning of images, as well as the generation of speech and the deeper meaning of speech, are processed together, increasing the complexity of processing.

[0063] To solve the above problems, this application provides a solution that clearly distinguishes the recognition of images and speech themselves from the recognition of the hidden meanings behind images and speech, thereby providing a method with a small computational workload that can be implemented using pure software programming, enabling a robot to have a certain human personality and a certain degree of thinking ability. This application abandons the complex processing process of CNNs and the way of swallowing everything whole by LLMs. Although this application uses language simulation, it does not involve language processing. Although the output result can be expressed in language, it can also be other non-verbal body movements.

[0064] After introducing the basic principle of this application, the following will specifically introduce various non-limiting embodiments of this application with reference to the accompanying drawings.

[0065] Exemplary Method

[0066] Figure 1 It is a schematic flowchart of a robot design method with a human personality and self-thinking ability provided by an embodiment of this application. As Figure 1 shown, this method includes the following content.

[0067] Step S110, obtain the five-sense information matrix;

[0068] Among them, the five-sense information matrix includes: image recognition information, sound recognition information, taste recognition information, olfactory recognition information, body touch and limb movement recognition information; the five-sense information matrix is a partial matrix inside the state matrix;

[0069] The information about the surrounding environment collected by the robot through other systems corresponds to the five human senses, namely vision, hearing, smell, taste, and touch. Specifically:

[0070] This information is integrated into a "five-sense information matrix", providing the robot with a comprehensive framework for perceiving the external environment.

[0071] In the solution provided by this application, the robot does not collect and generate the five-sense information matrix, but directly obtains it.

[0072] Among them, the five-sense information matrix includes: image recognition information, sound recognition information, taste recognition information, smell recognition information, body touch and limb movement recognition information;

[0073] The five-sense information matrix is the key for the robot to perceive the world. It integrates information from multiple aspects and provides a basis for the robot to understand the environment and make decisions. Among them, the acquisition methods of image recognition information include: the camera carried by the robot captures the light in the environment, converts it into digital signals, and analyzes and processes these signals through image recognition algorithms to extract the features in the images. Application examples are as follows: identifying the shape, color, size, position of objects, and facial features of people, etc. For example, the robot determines through image recognition that there is a table in front with a red cup on it, or recognizes that the approaching person is the owner. The acquisition methods of sound recognition information include: relying on a microphone to collect sound waves in the environment, converting them into electrical signals, and then using sound recognition technology to analyze the electrical signals to identify features such as the frequency, timbre, and loudness of the sound. Application examples are as follows: recognizing voice commands, such as when the owner says "turn on the TV", the robot can understand and execute the corresponding operation; it can also recognize the sounds in the surrounding environment, such as the honking of a car or the sound of rain, so as to judge the environmental conditions. The acquisition methods of taste recognition information include: with the help of taste sensors, by having a chemical reaction with the substances in contact, generating detectable signal changes to analyze the chemical composition and properties of the substances. Application examples are as follows: in some special scenarios, such as the food detection field, the robot can use taste recognition to judge whether the food has gone bad, or analyze the composition and taste characteristics of the food. The acquisition methods of olfactory recognition information include: using olfactory sensors, by sensing chemical substances in the air, interacting with the sensitive materials on the surface of the sensors to generate electrical signals, and then identifying the types and concentrations of odors. Application examples are as follows: recognizing whether there is a gas leakage smell in the kitchen, or helping to detect odors related to specific diseases in a hospital scenario to assist medical diagnosis. The acquisition methods of body touch recognition information include: touch sensors distributed on the surface of the robot's body. When stimulated by pressure, touch, vibration, etc., the physical properties of the sensors change, and these changes are converted into electrical signals to perceive touch information. Application examples are as follows: when the robot picks up an object, it can perceive the surface texture and weight of the object through touch recognition and avoid dropping the object due to improper grasping force; when contacting a person, it can perceive the contact force and position. The acquisition methods of limb movement recognition information include: through built-in motion sensors, such as accelerometers, gyroscopes, etc., monitoring information such as the angles, movement directions, and speeds of the robot's own joints to identify limb movements. Application examples are as follows: judging whether the robot itself is walking, running, grasping an object, or performing other complex action combinations, so as to adjust and optimize the actions to ensure the accuracy and smoothness of the actions.

[0074] Step S120, determine a stimulation matrix based on the change of the state matrix over time and a pre-determined personality matrix;

[0075] Among them, the personality matrix is used to characterize the personality of the robot; among them, the stimulus matrix is used to characterize the changes in the five-sense information matrix in the state matrix and the stimuli obtained under the influence of the personality matrix.

[0076] Based on the dynamic changes of the five-sense information matrix over time and the pre-set personality matrix, a stimulus matrix is generated. The personality matrix determines the sensitivity of the robot to different information and the reaction tendency. Under this influence, the changes in the five-sense information matrix are transformed into a stimulus matrix.

[0077] Example: If the robot's personality matrix shows that it is more sensitive to unfamiliar environments, when there are changes in the five-sense information matrix that it has never seen before, such as entering a new room with flashing lights, the robot combines these changes with its own personality characteristics to generate a stimulus matrix related to content such as "uneasy stimuli brought by unfamiliar environments and flashing lights".

[0078] Step S130, adjust the state matrix by combining the stimulus matrix, the personality matrix, and the pre-set memory warehouse.

[0079] The state matrix is used to characterize the current state of the robot and the external environment; among them, the current state of the robot is generally represented by the self matrix in the state matrix, and the current state of the external environment is generally represented by the five-sense information matrix in the state matrix.

[0080] The memory warehouse is constructed based on the changes in the state matrix.

[0081] Combine the stimulus matrix, the personality matrix, and the pre-set memory warehouse with each other to adjust the state matrix in this way. The state matrix reflects the current various states of the robot, and the memory warehouse accumulates information based on the changes in the state matrix to help the robot refer to past experiences when making decisions.

[0082] Example: Suppose that in a previous experience of the robot, a malfunction occurred when it encountered flashing lights, and this record is stored in the memory warehouse. When facing the stimulus of flashing lights again, the information in the stimulus matrix, the personality matrix, and the memory warehouse act together, which may increase the "vigilance level" parameter in the state matrix and adjust the "operating state" parameter to a more cautious mode.

[0083] Step S140, based on the adjusted state matrix, determine the behavior that the robot needs to perform to obtain a behavior matrix.

[0084] Specifically, based on the adjusted state matrix, the robot judges the behavior it needs to execute, thereby generating a behavior matrix. The behavior matrix contains instructions such as the language, thoughts, or actions that the robot is about to make.

[0085] Example: If the state matrix indicates that the robot is currently in a state of high alert and has detected potential dangers in its surroundings, the behavior matrix that the robot may determine is "emit an alarm sound, retreat, and activate the defense mode", and these instructions will drive the robot's hardware to perform corresponding actions.

[0086] Step S150, adjust the personality matrix by combining the state matrix and the stimulus matrix;

[0087] Adjust the personality matrix by combining the information contained in the state matrix and the stimulus matrix. This enables the robot to gradually change its personality characteristics according to its actual experiences and current state, realizing the dynamic development of its personality.

[0088] Example: If the robot, after encountering stimuli such as flashing lights multiple times, adjusts its state matrix and performs corresponding actions and finds that no serious consequences have occurred, then under the combined influence of the state matrix and the stimulus matrix, the sensitivity to flashing lights in the personality matrix may decrease and become relatively more "calm".

[0089] In some embodiments, it further includes:

[0090] Analyze the change process of the state matrix with a relevant relationship to obtain an analysis result;

[0091] The relevant relationship includes a causal relationship;

[0092] Update the memory warehouse based on the analysis result.

[0093] Specifically, the updating of the memory warehouse includes:

[0094] Associate, strengthen, merge, and delete the historical event information in the memory warehouse.

[0095] This process involves in-depth analysis of the changes in the robot's state matrix and the update of the memory warehouse based on these analysis results. The following are the specific steps and explanations:

[0096] Analyze the relevant relationship of the state matrix change;

[0097] First, collect the change data of the state matrix over a period of time, and this data includes the changes in the robot's internal state and the external environment.

[0098] Relevant relationship identification: Analyze these changes to identify which external events or internal state changes have caused the change in the state matrix. This may involve complex logical reasoning and pattern recognition techniques. Change process analysis: By analyzing the change process of the state matrix, determine the order, time points, and influencing factors of the changes. This helps to understand the dynamic process of state changes.

[0099] Analysis result: Finally, a detailed analysis result is obtained, which describes the causal chain of the state matrix changes, that is, which factors lead to which changes and how these changes affect each other.

[0100] Update the memory repository based on the analysis result, specifically including:

[0101] Association: In the memory repository, associate the new event with the existing historical events. If the new event is similar to a certain or certain historical events in some features, then these historical events will be activated, so as to help the robot utilize past experience when processing new events.

[0102] Reinforcement: For those specific historical events that are highly relevant to the new event, enhance their weights in the memory repository. This means that if a historical event is very similar to the new event, then this historical event will have a greater influence in future decisions.

[0103] Merge: If the new event is the same or similar to some historical events in the core features but there are some non-critical differences, then these historical events can be merged with the new event to simplify the memory repository and reduce redundancy.

[0104] Delete: Delete those historical events that are exactly the same as the new event or are obsolete and useless from the memory repository to maintain the efficiency and relevance of the memory repository.

[0105] Through these steps, the memory repository is continuously updated and optimized, enabling the robot to better learn from experience and improve the adaptability and efficiency of its decision-making and behavior. This update mechanism based on the correlation analysis enables the robot to maintain the relevance and effectiveness of its behavior and decision-making in a changing environment.

[0106] In some embodiments, it further includes: allocating calculation examples based on the principle of giving priority to ensuring the generation of the behavior matrix by the robot.

[0107] During the operation of the entire robot system, the reasonable allocation of computing resources is crucial for ensuring that the robot can generate the behavior matrix in a timely and accurate manner. The following is the detailed method of allocating calculation examples based on this principle:

[0108] On the basis of ensuring the generation of the behavior matrix, allocate part of the computing power to execute the steps "Analyze the change process of the state matrix with relevant relationships to obtain an analysis result; update the memory repository based on the analysis result."

[0109] In some embodiments, the behavior includes: language, idea or action. That is: This behavior matrix includes:

[0110] Language: What the robot needs to say or the signal it emits.

[0111] Idea: The decision-making logic and thinking process of the robot.

[0112] Action: The physical actions that the robot needs to perform.

[0113] The behaviors of the robot include language, ideas, or actions, which are the main ways for the robot to interact with the environment and respond to external stimuli.

[0114] In some embodiments, it further includes: obtaining a personality description; generating the personality matrix based on the personality description.

[0115] Collect descriptive data about the robot's personality. This data can be sourced from a variety of channels, including but not limited to:

[0116] User input: The user may directly provide preferences or requirements regarding the robot's personality.

[0117] Expert definition: Psychologists or behavioral scientists may provide professional descriptions of personality traits.

[0118] Cultural background: Consider the cultural background in which the robot will operate to ensure that the personality traits are suitable for the local culture.

[0119] Historical data: Analyze the personality traits exhibited by the robot in historical interactions.

[0120] Classify the collected personality descriptions according to different personality dimensions, such as extraversion, agreeableness, conscientiousness, emotional stability, and openness, etc.

[0121] Generate a personality matrix based on the personality description:

[0122] Quantify personality traits: Convert the personality description into quantifiable data. For example, psychological test results or behavioral ratings can be used to quantify personality traits.

[0123] Construct a personality matrix: Create a matrix where each row represents a personality trait and each column represents different personality dimensions or trait values. For example:

[0124] Extraversion: A rating from 1 - 5, where 1 represents very introverted and 5 represents very extroverted.

[0125] Emotional stability: A rating from 1 - 5, where 1 represents very unstable and 5 represents very stable.

[0126] Parameter assignment: Assign values to each parameter in the personality matrix according to the quantified data. These values define the robot's behavioral tendencies in different situations.

[0127] Dynamic adjustment: The personality matrix is not static and can be dynamically adjusted based on the robot's interaction with the environment and user feedback. For example, if the user feedback indicates that the robot is too direct, the agreeableness score in the personality matrix can be appropriately adjusted.

[0128] Influence on behavior decision-making: The personality matrix directly affects the robot's behavior decision-making. For example, a robot with a high extraversion score may be more inclined to actively communicate with the user, while a robot with a high emotional stability score may show more calmness when facing stress.

[0129] In some embodiments, based on the adjusted state matrix, determine the behaviors that the robot needs to perform to obtain a behavior matrix, specifically including: if the internal parameters of the adjusted state matrix meet the preset requirements, execute the behavior corresponding to the preset requirements; determine the behavior matrix based on the behavior corresponding to the preset requirements.

[0130] The adjusted state matrix is the updated state information of the robot after receiving external stimuli and combining its own personality matrix and memory repository. It contains the current internal and external state information of the robot, such as: External environment: time, location, status of surrounding objects or people, etc. Internal state: the robot's emotions, physical state, mental state, etc. Stimulus-related state: perception, judgment, and preliminary processing results of the stimulus. The preset requirements are the behavior trigger conditions predefined in the robot system. These conditions are usually related to certain parameters in the state matrix and are used to determine whether the robot needs to perform a specific behavior. For example: If the parameter representing "degree of danger" in the state matrix exceeds a certain threshold, trigger the "evasion" behavior. If the parameter representing "hunger level" in the state matrix reaches a certain threshold, trigger the "eating" behavior. If the parameter representing "emotional pleasure" in the state matrix reaches a certain threshold, trigger the "smiling" or "friendly interaction" behavior.

[0131] Judge whether the state matrix meets the preset requirements; the robot system will check each parameter in the adjusted state matrix to determine whether they meet the preset requirements. This step usually involves the following: Parameter threshold judgment: Check whether the key parameters in the state matrix exceed or are lower than the preset thresholds. Logical combination judgment: Multiple parameters may need to meet certain conditions simultaneously. For example, when "degree of danger is high" and "distance is close", trigger the "emergency evasion" behavior. Priority judgment: If multiple parameters meet different preset requirements simultaneously, the system needs to decide which behavior to execute according to the priority.

[0132] Once the parameters in the state matrix meet a certain preset requirement, the robot will execute the behavior corresponding to that requirement. For example: If the "degree of danger" parameter in the state matrix exceeds the threshold, the robot executes the "evasion" behavior. If the "emotional pleasure" parameter in the state matrix reaches the threshold, the robot executes the "smile" behavior.

[0133] The behavior matrix is a specific set of actions or instructions used to guide the robot to execute specific behaviors. The content of the behavior matrix depends on the matching result of the state matrix and the preset requirements. For example:

[0134] If the state matrix triggers the "evasion" behavior, the behavior matrix may include: Actions: Quickly retreat, change direction. Language: Emit a warning sound. Other instructions: Adjust the sensor scanning frequency to better monitor the surrounding environment. If the state matrix triggers the "smile" behavior, the behavior matrix may include: Actions: Facial muscle movement to form a smiling expression. Language: Say friendly words such as "Hello, nice to meet you".

[0135] After the behavior matrix is generated, the robot will send these instructions to specific execution components (such as motors, sensors, speech synthesizers, etc.) to complete the corresponding actions or behaviors.

[0136] Suppose the robot is in a certain scenario, and the state matrix is: S = [Time: 3 pm, Location: Park, Mood: Pleased, Hunger level: Medium, Danger level: Low, People around smiling]. The preset requirements include: If the "hunger level" reaches "high", then execute the "eating" behavior. If "emotional pleasure" and "people around smiling", then execute the "smiling response" behavior. In this example: The "emotional pleasure" and "people around smiling" in the state matrix meet the preset requirements of the "smiling response". Therefore, the robot executes the "smiling response" behavior, and the behavior matrix may include: Actions: Smile, wave. Language: Say "Hello, nice to meet you". In this way, the robot can make behavior decisions that conform to its "personality" and "situation" based on its own state and changes in the external environment.

[0137] In some embodiments, it further includes:

[0138] Automatically update the ego matrix in the state matrix over time; where the ego matrix is the parameter in the state matrix used to characterize the robot's state and changes over time.

[0139] Specifically, the "ME matrix" mentioned in the patent document is an important part of the state matrix (S), which is used to characterize the state of the robot itself and will change dynamically over time. This dynamic update mechanism is an important technical means to enable the robot to have human-like behaviors and adaptability. The following are the specific contents and steps for the automatic update of the "ME matrix" over time: The ME matrix is the core part of the robot's self-awareness. It contains various state information of the robot itself, such as: Physical state: such as battery level, health status, fatigue level, etc. Psychological state: such as emotions, motivations, interests, etc. Social state: such as identity, role, relationship with others, etc. Environmental adaptation state: such as familiarity with the current environment, task completion progress, etc. These state information are the basis for the robot to make decisions and behaviors. Therefore, keeping it updated is crucial for the intelligent behaviors of the robot.

[0140] The automatic update mechanism of my matrix ensures that the robot can dynamically adjust its own state according to the passage of time and environmental changes, so as to better adapt to the environment and make reasonable behaviors. The update mechanism generally includes the following aspects: Time-related updates; Periodic updates: At regular time intervals (such as every minute, every hour), my matrix will be updated according to preset rules. For example, the fatigue level of the robot may increase over time unless it performs rest or charging behaviors. Event-triggered updates: When a specific event occurs, my matrix will be updated immediately. For example, after the robot completes a task, its "task progress" status will be updated to "completed", and at the same time, its psychological states such as "sense of achievement" or "fatigue level" may be updated. Environment-related updates; External environment perception: The robot senses changes in the external environment through sensors and updates my matrix accordingly. For example, when the robot enters a new environment, its "environmental familiarity" status will decrease, and at the same time, its "alertness" status may increase. Interaction with others: When the robot interacts with humans or other robots, it will update my matrix according to the results of the interaction. For example, if the robot receives praise, its "mood" status will improve, while if it receives criticism, its "self-confidence" status may decrease. Behavior feedback updates; Behavior result evaluation: After the robot executes a behavior, it will update my matrix according to the feedback of the behavior result. For example, if the robot successfully completes a complex task, its "skill level" status will improve; if it fails, its "learning need" status may increase. Long-term goal adjustment: The robot will adjust my matrix according to the progress of long-term goals. For example, if the robot sets a long-term goal (such as learning a new skill), it will update its "learning progress" status according to the learning progress and adjust its "learning motivation" status as needed. The update of my matrix is usually based on preset rules and algorithms, which can be customized according to the design goals and application scenarios of the robot. For example: Time-based decay function: Some states (such as fatigue level) may gradually increase or decrease over time, and this change can be simulated through a decay function. Event-based increment function: When a specific event occurs, some states will be incrementally updated according to the nature and intensity of the event. For example, the amplitude of mood improvement when receiving praise can be determined according to the degree of praise. Behavior result-based feedback function: The behavior result will affect the state of the robot, and the state can be adjusted through a feedback function. For example, after successfully completing a task, the amplitude of skill level improvement can be determined according to the difficulty of the task.

[0141] This automatic update mechanism is of great significance in practical applications. For example: Home service robots can automatically adjust their states according to the passage of time and the activities of family members, so as to better serve family members. For example, they automatically enter the "quiet mode" at night to reduce noise; and enter the "active mode" in the morning to prepare for the activities of family members. For example, their battery levels change over time, etc.

[0142] Next, combining the above various preferred embodiments and referring to Figures 2-6 , the solution provided by this application will be further described:

[0143] In this solution, the result of the simulated robot having human personality characteristics is that when there is an external stimulus, the robot will make a human-like response to the stimulus. For example, when stimulated by negative language, it will show an angry expression, tone, and actions, etc.;

[0144] The solution provided by this application involves four important matrices:

[0145] "State matrix" S: This matrix is a set of parameters of various states of a robot at a certain moment, and each parameter represents a characteristic. Its value determines what the robot does at a certain moment and how it responds to stimuli. The parameters of the state matrix include external variables such as the time, location, and position of the robot, as well as the physical state parameters and mental consciousness state parameters of the robot itself. The parameters can be numbers or words. This matrix is variable and changes with time and the state of the robot's body. The characteristics represented by these parameters can be related or unrelated. The size of this matrix can be very large and can be a one-dimensional, two-dimensional, or three-dimensional matrix according to the depth of imitation. A simple example is

[0146] S(t) = [2024.11.06 School ME(t) Sunny] represents the current time, location, and weather conditions.

[0147] Among them, ME(t) is a more complex self-state matrix that describes the current state of "I". The following is just a simple example.

[0148] Self-state matrix ME: The self-state matrix describes "I", that is, the robot, and some current states, including physical states, social states, and all states except "personality characteristics". For example

[0149] ME(t) = [Student, 20 years old, 175 cm tall, female, of medium build, from a wealthy family, having a stomachache, dreaming of being a singer, studying hard].

[0150] The ME matrix also changes with time.

[0151] Personality Matrix C: This matrix is a set of personality traits that a robot possesses. It determines how a robot responds to stimuli. The personality matrix is jointly determined by the robot's DNA and birth chart. Historical experiences such as injuries can change this personality matrix. Therefore, it is also variable, changing over long periods of time and also having minor changes within a day, influenced by the five elements.

[0152] The personality matrix consists of two important matrices: C1 and C2.

[0153] C1 is a matrix that describes MBTI personality traits and other characteristics. As shown in the following example.

[0154]

[0155] C2 is a matrix of the combination of the Chinese zodiac sign of the year and the constellation of the month. As shown below.

[0156]

[0157] C1 and C2 jointly determine a person's personality matrix C.

[0158] C = C2[i,j] * C1. Where i and j are the Chinese zodiac sign of the year and the constellation of the month corresponding to a certain robot respectively. In this way, by looking up the C2 table, a number is found as the weight of the C1 matrix. The above is just an example, and the effect of the C2 matrix on the C1 matrix can also be in other forms. For example, select some items in C2 for weighted multiplication.

[0159] The C1 and C2 matrices can have different modes of action, C = C2[i,j] * C1. Or let C2 change with time, adding time t, and there can also be the following relationship:

[0160] C(t) = C2[i,j](t) * C1(t) and integrate.

[0161] "Stimulus Matrix" T: This matrix represents stimulus characteristics in matrix form. It can change the values of the "State Matrix" according to the "Personality Matrix". It is similar to the "State Matrix" but has far fewer parameter items. The determination of some of its parameters requires comparison with some parameters in the ME matrix. Therefore, it can be divided into two parts T(t) = [T1(t) T2(t)].

[0162] Among them, T1 is stimulus information that has nothing to do with the robot, and T2 is stimulus information related to the robot.

[0163] An example of the T1 matrix:

[0164] T1 = [dark blue clothes, male, looking straight ahead, 3 meters away, no change in body, no change in voice, no change in smell, no change in taste, old, about 40 years old, fat, awkward, getting wet in the rain]

[0165] T2 = [There is ice cream I like in the hands of someone older than me. That person is not my competitor and our relationship is unknown. I don't know them.]

[0166] In some embodiments, it further includes: Behavior matrix E: This is the action matrix of various parts such as the body or language generated after stimulus processing.

[0167] For example: E = [smiling, raising hand, kicking leg]. This matrix will be input to the specific execution hardware to generate corresponding actions or language.

[0168] Specifically, the overall operation mode is as follows:

[0169] The overall operation mode of the robot's thinking is as Figure 2 shown. Briefly speaking, it is that the "stimulus matrix" determines how the "state feature matrix" changes under the action of the "personality matrix". If the change (which can be continuous increase, or change between 0 and 1) exceeds the threshold of a certain item, a response will be made to the stimulus. If it does not exceed the threshold, there will be no response. At the same time, this stimulus will also gradually change the value of the "personality matrix". As shown below.

[0170] S(t + 1) = f(C(t), S(t), T(t)).

[0171] C(t + 1) = g(C(t), S(t), T(t))

[0172] Among them, the parameter t represents the value at the current moment, and t + 1 represents the value at the next moment.

[0173] f and g represent two functions. That is to say, the value of the state feature matrix at the next moment is a function of the state feature matrix, personality matrix, and stimulus matrix at the current moment.

[0174] An example of the implementation of functions f and g is as follows:

[0175] S(t + 1) = f(C, S, T) = C(t) * S(t) + T(t).

[0176] C(t + 1) = g(C, S, T) = C(t) + S(t) + 0.1T

[0177] The implementation of personality can also be point-to-point, that is, a certain item in S represents a certain state and is a function of other parameters. Different items in S can have different functions. Assuming a three-dimensional S, one item can be expressed as, that is,

[0178] S(i, j, k) = f(S(x, y, z), C(l, m, n), T(a, b, c))

[0179] Where i, j, k, x, y, z, l, m, n, a, b, and c can be any number or a set of numbers. When it is a set, it means that one or more items in S are functions of one or more items in each matrix on the right side of the equation.

[0180] Refer to Figure 2 , the operation mode of the personality matrix includes:

[0181] The "personality matrix" can be described by thousands of personality characteristic parameters. These characteristic parameters are generated by larger personality characteristic parameters: such as the 4 personality types of MBTI [introverted / extroverted, perceiving / judging, feeling / thinking, sensing / intuitive]. Or twelve major personality characteristics similar to the twelve Chinese zodiac signs. [The personality of the dragon, the personality of the sheep, the personality of the rabbit, etc.]. These major personality characteristics can generate smaller-level personality characteristics, such as [slow to warm up / fast to warm up, sentimental / free and easy, sensitive and suspicious / trusting, calm / composed, etc.]. These personality characteristics can also generate even smaller-level personality characteristics.

[0182] These characteristics are determined by the robot's DNA and birth date. Because the year of birth determines the personality characteristics of the twelve Chinese zodiac signs, the month determines the smaller-level constellation personality characteristics, the day has 30 possible personality characteristics, the hour has 24 personality characteristics, the minute has 60 personality characteristics, and the second has 60 personality characteristics.

[0183] Historical experiences will change the parameters of this personality matrix. For example, being bitten by a dog once will make one sensitive to dogs, and the parameter value of "sensitive to dogs" will increase.

[0184] The personality characteristics of a robot at a certain moment are also related to the current time. That is to say, the innate personality characteristics and the five elements jointly determine the characteristics at the current moment.

[0185] C(t + 1) = h(C(t), time)

[0186] Where time is a matrix representing time, such as [July 12, 2025, 3:00 PM, 25 minutes, 30 seconds]

[0187] An example of the implementation of the h function. For example, if time is the above time matrix, then

[0188] The personality value representing "irritable" in the C matrix will increase because the weather is relatively hot in July and it is 3:00 PM.

[0189] Such as Figure 4 As shown, the process of comprehensive processing has the following steps:

[0190] Simulate the recognition functions of the five human senses. This patent assumes that the recognition function has been completed, so it does not involve the algorithms for how to recognize. With the current development of AI technology, it is already capable of correctly recognizing images, actions, colors, sounds, etc. This recognition will form conceptual parameters such as [dog, yellow, 35Hz]. This patent focuses on the processing of conceptual parameters after recognition. In the "Comprehensive Processing" module, the "five sense organs" perform the recognition of stimuli and form characteristic parameters.

[0191] For example: "A man wearing dark blue clothes, with his eyes looking straight ahead, waved his hand at a distance of 3 meters. It was raining, and there were no changes in other external environments, no changes in the robot's body, no sound changes, no smell detected, and no taste changes on the tongue."

[0192] This stimulus situation can be represented by a recognition matrix as: T1 = [dark blue clothes, male, eyes looking straight, 3 meters away, it's raining, no external environment changes, no body changes, no sound changes, no smell changes, no taste changes]

[0193] Form subjective concepts for the stimulus (based on the personality matrix and historical experience). For example: The dark blue clothes are very "old", the male is "about 40 years old", is very "fat", the movement is a bit "awkward", not holding an umbrella will "get wet in the rain", and there is a possibility of "no harm" to the robot's body. Based on the previous T, this feature is T1 = [old, about 40 years old, fat, awkward, get wet in the rain, no harm]

[0194] Whether it is related to "me": After comparing the aforementioned stimulus with "me", the "consciousness" module will form the following parameters: T2 = [older than me, not my competitor, relationship with me unknown]

[0195] "Me" has three choices of stimulus responses to decide how to respond to the stimulus: ego, id, and superego. This is the social attribute of "me". The "ego" is the current state shown, such as the state of an "excellent student". The "id" is the original state of me in the subconscious, such as "being playful but liking to draw". The "superego" is the "excellent" social state in the subconscious, such as "strictly demanding myself to be a better self".

[0196] Form a stimulus matrix. For example, T = [T1 T2] = [dark blue clothes, male, eyes looking straight, 3 meters away, no body changes, no sound changes, no smell changes, no taste changes, old, about 40 years old, fat, awkward, get wet in the rain, no harm, older than me, not my competitor, relationship with me unknown]

[0197] Act on the "state matrix" in combination with the "personality matrix". This is a rather complex process of action:

[0198] For example, the threshold of the "sensitive" item in the "personality matrix" is low, such as 0.1. One way to process it is to use 0.1*T to change the stimulus matrix.

[0199] Another example is that the threshold of the "caring" item in the "personality matrix" is high, assumed to be 1.2. Then, "getting caught in the rain" will trigger the "caring" to start working. As a result, the value of [holding an umbrella for him] in the status matrix S is updated to exceed the threshold.

[0200] Updating the status matrix: After the previous status matrix S is updated, corresponding operations will be executed according to whether the current working state is "ego", "id", or "superego". When these three start working is determined by the time, location, etc. in the personality matrix C and the status matrix S. For example, [at home at night] will show the "id". [at school during the day] will show the "ego". [on the stage at school] will show the "superego". If the value of [self-awareness] in the personality traits is relatively high, the "superego" will also be shown in the above three situations.

[0201] The process of reacting to stimuli: When the "status matrix" changes, the robot will make corresponding reactions according to the change of the threshold and the category of the status change parameters, including the reactions of various parts of the body, as well as the reactions of thoughts and language. The reaction can be a subconscious action similar to a conditioned reflex, or an action that requires thinking and calculation.

[0202] For example: When the item [holding an umbrella for him] in the status matrix exceeds the threshold, it will trigger a series of actions of the body and limbs. If the distance is far, it is necessary to first execute [walk over], and then execute a series of actions of [holding an umbrella for someone] in the predetermined program.

[0203] Furthermore, during the sleep process of the machine, since there are fewer stimuli during sleep, only a part of the computing power, A%, is used to process the stimuli. The remaining part, B%, will do "subconscious" work. Among them, A% + B% = 100%. The proportions of A and B in 100% will vary according to different tasks during sleep.

[0204] Generally, A = 50 and B = 50. That is, half of the computing power is used to process the stimuli, and half of the computing power is used for "garbage collection" work. It can also be A = 20 and B = 80. That is, 20% of the computing power is used to process the stimuli, and 80% of the computing power is used for "subconscious" work.

[0205] "Subconscious" is the information "warehouse" of the robot, used to store all the things it has experienced. The sleep process is a process of "garbage disposal" for this "warehouse". Every day, new stimuli enter the warehouse, so this "garbage" disposal work needs to be done every day.

[0206] The methods of processing the "subconscious" when sleeping are: association, reinforcement, merging, and deletion.

[0207] Association includes:

[0208] It occurs when the new stimulus is related to the content in the "warehouse". Through certain processing such as logical reasoning, random change, or deterministic change, the cognitive understanding of historical similar content can be updated. For example, the historical record is [Puppies are cute and can be petted], but if the experience of the day is [Puppies are cute but will bite when petted], the record will be updated to [Puppies are cute and can be petted, maybe they will bite].

[0209] Reinforcement includes: It occurs when the new stimulus is the same as the content in the "warehouse", but certain parameters determine that they cannot be merged. At this time, the weight of this content in the warehouse will increase.

[0210] For example: [Puppies are cute, can be petted, will bite and it hurts]. If this situation occurs again for the second time, because [pain] is a key parameter, this item will not be merged, but the value of the item "pain" will be strengthened.

[0211] Merger includes:

[0212] It occurs when the new stimulus is the same as the content in the "warehouse", and there are no parameters determining that reinforcement is needed, then the content is merged. For example

[0213] In the warehouse there is: [The path is quiet, with fallen leaves, autumn is serene]. If one day there appears [The path is quiet, with fallen leaves, autumn is serene, and a deer rushes out]. Then the previous one in the warehouse can be deleted and updated to [The path is quiet, with fallen leaves, autumn is serene, and a deer rushes out].

[0214] Deletion includes:

[0215] It occurs when the new stimulus is exactly the same as the content in the "warehouse", and there are no parameters determining that reinforcement is needed, then the content is deleted. Clean up the memory. For example, if the same path has been walked many times, only one record is needed.

[0216] These are the four main operations. In some embodiments, other garbage disposal operations can also be arranged.

[0217] In some embodiments, the following characteristic parameter values are defaulted to 1.

[0218] The initial "state matrix" is S = [Sunny, school dismissal time, in a good mood, having nothing to do],

[0219] The "personality matrix" is C = [Loving animals, not sensitive, kind-hearted, emotional],

[0220] The "stimulus matrix" is T = [Puppy, animal moving, wagging tail, smiling, normal heart rate]

[0221] The reaction made by the robot will be: Modify the "state matrix" according to the "personality matrix" to:

[0222] S = [Sunny, time to go home from school, happy mood, nothing to do, walking towards the animals]

[0223] The state matrix of this change causes the robot's body to move with E = [walk forward, smile, slow speed]

[0224] At this time, the "state feature" matrix is: S = [sunny, school time, happy mood, nothing to do, smiling, moving forward, slow speed]

[0225] When you walk in front of the puppy, the "state matrix" is modified to:

[0226] S = [After school, feeling happy, nothing to do, animals are close by]

[0227] The “stimulus matrix” is T = [puppy, animal moving, wagging tail, smiling, normal heartbeat]

[0228] The robot will modify the "state matrix" according to [loving animals, insensitive] in the "personality matrix" to: S = [sunny, time to go home from school, happy, nothing to do, smiling, still, squatting]

[0229] The change of the state matrix causes the actuator to perform the action: E = [squat].

[0230] The movement involves not only bending the legs, but also bending the body slightly forward.

[0231] If at this time, the "stimulus matrix" becomes T = [the puppy moves, the speed is not fast, the heartbeat is normal, and the distance is getting farther and farther]

[0232] According to the characteristics of the "moving" object, check the "personality matrix" to see that the sensitive item you want to associate is [insensitive], and update the "state matrix" to:

[0233] S = [Sunny, time to go home from school, happy mood, nothing to do, smiling, still, squatting, eyes fixed on the puppy].

[0234] The process of updating the state matrix from the personality matrix is ​​relatively complicated. This patent is just an example, and there are many other ways.

[0235] In some embodiments:

[0236] State feature matrix S = [night, 11pm, alley, female, 20 years old, not ugly, on the way home]

[0237] Stimulus matrix T1 = [dark blue clothes, male, eyes looking straight ahead, 3 meters away, raining, no changes in external environment, no changes in body, no changes in sound, no changes in smell, no changes in taste]

[0238] Formed after preliminary processing, the updated stimulus matrix T1 = [old, about 40 years old, fat, awkward, getting caught in the rain, harmless]

[0239] "ME" in the personality matrix = [female, student, 20 years old, long hair, oval face, standard weight, no disease, good at painting, likes listening to music, owns an iPhone, excellent grades, etc.]

[0240] After comparison with "ME", the following parameter matrix T2 = [older than me, not my competitor, relationship with me unknown, etc.] will be formed

[0241] Form the updated stimulus matrix:

[0242] T = [T1 T2] = [dark blue clothes, male, looking straight ahead, 3 meters away, no change in body, no change in voice, no change in smell, no change in taste, old, about 40 years old, fat, awkward, getting caught in the rain, harmless, older than me, not my competitor, relationship with me unknown]

[0243] If there is an item [sensitive] in the "personality trait" matrix, a [dodge] command will be formed to update the "status trait matrix" to:

[0244] S = [night, 11 pm, alley, female, 20 years old, not ugly, on the way home, running fast, away from the direction of the man across the street]

[0245] Thus, the "running" mode is driven to start. If it is detected that the man is running towards her, update the "status matrix" to S = [night, 11 pm, alley, female, 20 years old, not ugly, on the way home, running fast, away from the direction of the man across the street, accelerating] and at the same time generate the behavior matrix E = [running fast, away from the direction of the man across the street, accelerating]

[0246] Exemplary Apparatus

[0247] The device embodiment of the present application can be used to execute the method embodiment of the present application. For details not disclosed in the device embodiment of the present application, please refer to the method embodiment of the present application.

[0248] Figure 7 The block diagram of a robot design device with self - thinking human personality provided by an embodiment of the present application is shown as follows. As Figure 7 shown, the device includes:

[0249] An acquisition module 71, configured to acquire a five - sense information matrix;

[0250] Wherein, the five - sense information matrix includes: image recognition information, sound recognition information, taste recognition information, smell recognition information, body touch and limb movement recognition information;

[0251] The first determination module 72 is configured to determine a stimulus matrix based on the change of the five-sense information matrix over time and a pre-determined personality matrix;

[0252] Wherein, the personality matrix is used to characterize the personality of the robot;

[0253] Wherein, the stimulus matrix is used to characterize the stimulus obtained under the influence of the personality matrix from the change of the five-sense information matrix;

[0254] The combination module 73 is configured to adjust the state matrix by combining the stimulus matrix, the personality matrix, and a preset memory repository;

[0255] The state matrix is used to characterize the current state of the robot;

[0256] The second determination module 74 determines the actions that the robot needs to perform based on the adjusted state matrix, and obtains an action matrix;

[0257] The adjustment module 75 is configured to adjust the personality matrix by combining the state matrix and the stimulus matrix;

[0258] The memory repository is constructed based on the change of the state matrix.

[0259] Exemplary Electronic Device

[0260] Next, refer to Figure 8 to describe an electronic device according to an embodiment of the present application. Figure 8 The block diagram of the electronic device according to the embodiment of the present application is illustrated.

[0261] As Figure 8 shown, the electronic device 800 includes one or more processors 810 and a memory 820.

[0262] The processor 810 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 800 to perform desired functions.

[0263] The memory 820 may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage media, and the processor 810 may run the program instructions to implement the method for designing a robot with a human-like personality and self-thinking ability according to various embodiments of the present application described above, as well as / or other desired functions. Various contents such as category correspondence relationships may also be stored in the computer-readable storage media.

[0264] In one example, the electronic device 800 may further include: an input device 830 and an output device 840, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown).

[0265] In addition, the input device 830 may further include, for example, a keyboard, a mouse, an interface, etc. The output device 840 may output various information to the outside, including analysis results, etc. The output device 840 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0266] Of course, for simplicity, Figure 8 only some of the components related to the present application in the electronic device are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, according to specific application scenarios, the electronic device may further include any other appropriate components.

[0267] Exemplary Computer Program Product and Computer Readable Storage Medium

[0268] In addition to the above methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions that, when run by a processor, cause the processor to execute the steps in the method for designing a robot with a human-like personality and self-thinking ability according to various embodiments of the present application described in the "Exemplary Methods" section above.

[0269] The computer program product may be written in any combination of one or more programming languages for executing the program code of the operations of the embodiments of the present application. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, executed as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0270] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, and when the computer program instructions are run by a processor, the processor is caused to execute the steps in the method for designing a robot with human-like personality and self-thinking according to various embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0271] The computer-readable storage medium may employ any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium may, for example, include but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0272] The above description has been given for purposes of illustration and description. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions, and sub-combinations thereof.

Claims

1. A method for designing a robot with human characteristics and self-thinking, characterized in that: include: Obtain the five senses information matrix; The five sense information matrix includes: image recognition information, sound recognition information, taste recognition information, smell recognition information, body touch and body movement recognition information; the five sense information matrix is ​​a partial matrix inside the state matrix; determining a stimulus matrix based on the change of the state matrix over time and a predetermined personality matrix; Wherein, the personality matrix is ​​used to characterize the personality of the robot; Wherein, the stimulus matrix is ​​used to characterize the changes of the state matrix, and the stimulus obtained under the influence of the personality matrix; Combining the stimulus matrix, the personality matrix and the preset memory warehouse, adjusting the state matrix; The state matrix is ​​used to represent the current state of the robot and the external environment; Based on the adjusted state matrix, determine the behavior that the robot needs to perform and obtain the behavior matrix; combining the state matrix and the stimulus matrix to adjust the personality matrix; The memory warehouse is constructed based on changes in the state matrix.

2. The method for designing a robot with human personality and self-thinking according to claim 1, characterized in that: Also includes: Analyze the change process of the state matrix with correlation relationship and obtain the analysis result; The correlation relationship includes a causal relationship; Based on the analysis result, the memory warehouse is updated.

3. The method for designing a robot with human character and self-thinking according to claim 2, characterized in that: Updating the memory warehouse includes: The historical event information in the memory warehouse is associated, strengthened, merged and deleted.

4. The method for designing a robot with human character and self-thinking according to claim 3, characterized in that: Also includes: Based on the principle of giving priority to ensuring that the robot generates the behavior matrix, the examples are allocated.

5. The method for designing a robot with human character and self-thinking according to claim 1, characterized in that: The behavior includes: words, thoughts or actions.

6. The method for designing a robot with human personality and self-thinking according to claim 1, characterized in that: Also includes: Get personality descriptions; Based on the personality description, the personality matrix is ​​generated.

7. The method for designing a robot with human character and self-thinking according to claim 1, characterized in that: Based on the adjusted state matrix, determine the behavior that the robot needs to perform and obtain the behavior matrix, including: If the adjusted internal parameters of the state matrix meet the preset requirements, the behavior corresponding to the preset requirements is executed; A behavior matrix is ​​determined based on the behavior corresponding to the preset requirement.

8. The method for designing a robot with human personality and self-thinking according to claim 1, characterized in that: Also includes: automatically updating the I-matrix in the state matrix over time; The I-matrix is ​​a parameter in the state matrix that is used to characterize the state of the robot and changes with time.

9. A robot design device with human characteristics and self-thinking, characterized in that: include: An acquisition module, used to acquire the five senses information matrix; The five sense information matrix includes: image recognition information, sound recognition information, taste recognition information, smell recognition information, body touch and body movement recognition information; the five sense information matrix is ​​a partial matrix inside the state matrix; A first determination module, configured to determine a stimulus matrix based on changes in the state matrix over time and a predetermined personality matrix; Wherein, the personality matrix is ​​used to characterize the personality of the robot; Wherein, the stimulus matrix is ​​used to characterize the changes of the state matrix, and the stimulus obtained under the influence of the personality matrix; A combining module, used for combining the stimulus matrix, the personality matrix and the preset memory warehouse to adjust the state matrix; The state matrix is ​​used to represent the current state of the robot and the external environment; A second determination module is used to determine the behavior that the robot needs to perform based on the adjusted state matrix to obtain a behavior matrix; An adjustment module, used for adjusting the personality matrix by combining the state matrix and the stimulus matrix; The memory warehouse is constructed based on changes in the state matrix.

10. An electronic device, characterized in that: include: A processor, and a memory for storing a program executable by the processor; The processor is used to implement the method for designing a robot with human personality and capable of self-thinking as described in any one of claims 1 to 8 by running the program in the memory.

Images