CN107633292B - Robot emotion space modeling method based on quantum structure - Google Patents

Robot emotion space modeling method based on quantum structure Download PDF

Info

Publication number
CN107633292B
CN107633292B CN201710902831.1A CN201710902831A CN107633292B CN 107633292 B CN107633292 B CN 107633292B CN 201710902831 A CN201710902831 A CN 201710902831A CN 107633292 B CN107633292 B CN 107633292B
Authority
CN
China
Prior art keywords
quantum
emotion
space
state
time
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201710902831.1A
Other languages
Chinese (zh)
Other versions
CN107633292A (en
Inventor
闫飞
焦思皓
陈克寒
蒋振刚
杨华民
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changchun University of Science and Technology
Original Assignee
Changchun University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Changchun University of Science and Technology filed Critical Changchun University of Science and Technology
Priority to CN201710902831.1A priority Critical patent/CN107633292B/en
Publication of CN107633292A publication Critical patent/CN107633292A/en
Application granted granted Critical
Publication of CN107633292B publication Critical patent/CN107633292B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Feedback Control In General (AREA)
  • Manipulator (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

The invention relates to a quantum structure-based robot emotion space modeling method, which is characterized by comprising the following steps of: in a quantum system, a quantum psychological judgment space is constructed on the basis of a classical psychological judgment space; the quantum psychology judgment space is a two-dimensional coordinate system, the coordinate axes respectively represent excitement and happiness, and each emotion can be quantized into a corresponding value in the coordinate system; the two-dimensional representation provides a visual method for observing general features between different emotions, and particularly, the emotion of the robot is more direct compared with the subtle emotion of a human. Two quantum sequences are used for determining a specific emotion in a psychological judgment space, and the other quantum sequence is used for determining time information corresponding to the emotion. Then, as time changes, the emotions of the robot at different moments are added to the emotion space one by one. And finally, manipulating the quantum emotion space by using various quantum logic gates, including emotion space initialization, emotion transformation and emotion retrieval.

Description

Robot emotion space modeling method based on quantum structure
Technical Field
The invention relates to a quantum structure-based robot emotion space modeling method, and belongs to the field of artificial intelligence.
Background
The emotion space is an effective method for describing the emotional states of the robot or the human at different times and ensuring good interaction between the human and the robot, and is an important research direction in the field of artificial intelligence. In recent years, along with the development of quantum computing and quantum information, the robot emotion space based on a quantum structure is also receiving wide attention, and the parallelism and the reversibility of quantum mechanics enable quantum information processing to show great advantages in relevant applications. The psychological judgment space is a commonly used method in the emotion data visualization process, and corresponding points can be found in a two-dimensional coordinate system for each emotion. The robot emotion representation method is expanded to the field of quantum computing, and a novel robot emotion space modeling method based on a quantum structure is realized by researching and expanding a psychological judgment space and reasonably applying various quantum gates.
Disclosure of Invention
The invention aims to provide a robot emotion space modeling method based on a quantum structure. Firstly, a three-dimensional quantum emotion space for describing the emotion change of the robot is constructed, and an emotion point is positioned in the three-dimensional emotion space by utilizing the tensor product of three quantum sequences, wherein two quantum sequences are used for determining a specific emotion in a psychological judgment space, and the other quantum sequence is used for determining time information corresponding to the emotion. Then, as time changes, the emotions of the robot at different moments are added to the emotion space one by one. And finally, manipulating the quantum emotion space by using various quantum logic gates, including emotion space initialization, emotion transformation and emotion retrieval.
The technical scheme of the invention is realized as follows: a robot emotion space modeling method based on a quantum structure is characterized by comprising the following steps: firstly, in a quantum system, constructing a quantum psychological judgment space on the basis of a classical psychological judgment space; the quantum psychology judgment space is a two-dimensional coordinate system, the coordinate axes respectively represent excitement and happiness, and each emotion can be quantized into a corresponding value in the coordinate system; the two-dimensional representation provides a visual method for observing general features between different emotions, and particularly, the emotion of the robot is more direct compared with the subtle emotion of a human. The quantum psychology judgment space uses 2n quantum bits to represent position information of the emotion in a coordinate system, wherein the first n quantum bits are used for representing information of the emotion along an excitation coordinate axis, and the last n quantum bits are used for representing information of the emotion along a happy coordinate axis; quantum emotion space, QES, is composed of Quantum psychological judgment space and a time axis, the time axis is represented by m Quantum bits, therefore, we use 2n + m Quantum bits to integrate QES information into a normalized Quantum state, i.e. Quantum emotion space expression | e (t) > can be defined as:
Figure BDA0001423413820000011
wherein, | t>Representing the time information of each emotion in QES, | At>And | Pt>Respectively represent | t>The position information of the time emotion in the excitation coordinate axis and the happy coordinate axis is defined as:
Figure BDA0001423413820000012
Figure BDA0001423413820000013
wherein the content of the first and second substances,
Figure BDA0001423413820000014
and
Figure BDA0001423413820000015
in a substantially quantum state, i.e.
Figure BDA0001423413820000016
The method comprises the following steps:
step 1, preparing 2n + m quantum bits in quantum system, its state is 2n + m |0>Is the tensor product of
Figure BDA0001423413820000017
We perform Hadamard transform, H transform, on m qubits representing temporal information, the H transform being a unitary matrix expressed by
Figure BDA0001423413820000018
Separately H transforming each of the m qubits may yield 2mSuperposition of individual quantum states, i.e. producing from 0 to 2m-1 all binary values, each elementary state existing simultaneously with a probability of being present
Figure BDA0001423413820000021
In view of the quantum system integrity, in performing the H transform, we process other 2n qubits using a special two-dimensional identity matrix I, where I is expressed as
Figure BDA0001423413820000022
By conversion, the time period of the robot emotion space is initialized, namely, the time period is changed
Figure BDA0001423413820000023
Is converted into
Figure BDA0001423413820000024
Step 2, according to the conversion result of step 1, the quantum system at this time comprises 2mEach basic state is composed of 2n quantum bits representing emotional information and unique time information, no emotional state exists in the emotional space of the robot, and when the robot receives an instant stimulus from the outside, the quantum system converts the stimulus into a conversion matrix M which can be identified by the emotional space1Since the fundamental state in the stable quantum system is in the entangled state, we need to use special quantum operations to ensure that the other 2 s are in the process of initializing emotionmThe-1 fundamental state is not affected. In quantum computing, the outer product | a><a | pair of entangled quantum states | a |>Projection, i.e. at | a>The quantum state is measured in the direction, therefore, with |1><1| Pair of entangled quantum states |1>Projection is obtained
Figure BDA0001423413820000025
Simultaneous calculation
Figure BDA0001423413820000026
The emotional state at the time t 1 can be obtained. Then use
Figure BDA0001423413820000027
Making entangled quantum states respectively pair i>And projecting to obtain the basic state at other moments.
Step 3, the robot receives various stimulations from the outside due to the constant change of the external environment, so that the emotional state of the robot has continuity and time phaseAnd (4) turning off. The emotional state of the robot at the time t is related to the received stimulus and the emotional state at the time t-1; using transformation matrix M1,M2,M3,...,MtObtaining an emotion conversion matrix phi at the moment tt＝M1M2M3...MtThen calculate
Figure BDA0001423413820000028
The emotional state at time t can be obtained. Use 2m1 emotion conversion matrix can initialize a unique emotion state at each time in the quantum emotion space.
Step 4, utilizing quantum computation and projection measurement commonly used in quantum information to carry out entanglement state 2 in quantum emotion spacemThe measurement is carried out on each basic state, and each measurement can be carried out on information from one basic state in the quantum emotion space, wherein the information comprises time information and the emotional state at the corresponding moment. When all the information is retrieved, the original quantum emotion space can be obtained.
The invention has the following positive effects: the robot emotion space in the field of artificial intelligence is expanded to the field of quantum computing, and the parallelism in the quantum computing is utilized, so that fewer resources are utilized in the construction and initialization processes of the emotion space. Quantum mechanics ensures that all transformations are invertible, and the proposed emotion transformation matrix can more easily track emotion changes. With the continuous breakthrough of research in the fields of quantum computing and quantum information, a quantum robot equipped with a quantum computer and an auxiliary system can be expected to perform special tasks in a continuously changing external environment. The method makes up the blank of the emotion space field of the quantum robot, and provides a solid foundation for the subsequent research of the quantum robot.
Drawings
Fig. 1 is a quantum psychology decision space represented by 3 qubits.
Fig. 2 shows a quantum emotion space for initializing an emotion state at time t-1.
FIG. 3 is a quantum emotion space for initializing all moments' emotional states.
Detailed Description
The invention is further described with reference to the accompanying drawings in which: a robot emotion space modeling method based on a quantum structure is characterized by comprising the following steps: firstly, in a quantum system, a quantum psychological judgment space is constructed on the basis of a classical psychological judgment space. The quantum psychology judgment space is a two-dimensional coordinate system, the coordinate axes respectively represent excitement and happiness, and each emotion can be quantized into a corresponding value in the coordinate system. The two-dimensional representation provides a visual method for observing general features between different emotions, and particularly, the emotion of the robot is more direct compared with the subtle emotion of a human. The quantum psychology decision space uses 2n quantum bits to represent the position information of emotion in the coordinate system (the first n quantum bits are used to represent the information of emotion along the "excitement" axis, and the last n quantum bits are used to represent the information of emotion along the "happy" axis, as shown in fig. 1). Quantum Emotion Space (QES) is composed of a Quantum psychological decision space and a time axis, which is expressed using m Quantum bits. Therefore, we use 2n + m quantum bits to integrate the QES information into one normalized quantum state, i.e., the quantum emotion space expression | e (t) > can be defined as:
Figure BDA0001423413820000029
wherein, | t>Representing the time information of each emotion in QES, | At>And | Pt>Respectively represent | t>The position information of the time emotion in the excitation coordinate axis and the happy coordinate axis is defined as:
Figure BDA0001423413820000031
Figure BDA0001423413820000032
wherein the content of the first and second substances,
Figure BDA0001423413820000033
and
Figure BDA0001423413820000034
in a substantially quantum state, i.e.
Figure BDA0001423413820000035
The method comprises the following steps:
step 1, preparing 2n + m quantum bits in quantum system, its state is 2n + m |0>Is the tensor product of
Figure BDA0001423413820000036
We perform a Hadamard transform, i.e. H-transform, on the m qubits representing the temporal information. H transform is a unitary matrix expressed as
Figure BDA0001423413820000037
Separately H transforming each of the m qubits may yield 2mSuperposition of individual quantum states, i.e. producing from 0 to 2m-1 all binary values, each elementary state existing simultaneously with a probability of being present
Figure BDA0001423413820000038
In view of the quantum system integrity, in performing the H transform, we process other 2n qubits using a special two-dimensional identity matrix I, where I is expressed as
Figure BDA0001423413820000039
By conversion, the time period of the robot emotion space is initialized, namely, the time period is changed
Figure BDA00014234138200000310
Is converted into
Figure BDA00014234138200000311
Step 2, conversion according to step 1Alternatively, the quantum system at this time comprises 2mA basic state. Each elementary state consists of 2n qubits representing affective information and a unique piece of time information. At this time, there is no emotional state in the robot emotional space. When the robot receives an instant stimulus from the outside, the quantum system converts the stimulus into a conversion matrix M which can be identified by emotion space1. Since the fundamental state in the stable quantum system is in the entangled state, we need to use special quantum operations to ensure that other 2 s are in the process of initializing emotionmThe-1 fundamental state is not affected. In quantum computing, the outer product | a><a | pair of entangled quantum states | a |>Projection, i.e. at | a>The quantum state is measured in the direction, therefore, with |1><1| Pair of entangled quantum states |1>Projection is obtained
Figure BDA00014234138200000312
Simultaneous calculation
Figure BDA00014234138200000313
The emotional state at the time t 1 can be obtained. Then use
Figure BDA00014234138200000314
Making entangled quantum states respectively pair i>And projecting to obtain the basic state at other moments.
And 3, the robot receives various stimulations from the outside due to the constant change of the external environment. Therefore, the emotional state of the robot has continuity and time correlation. The emotional state of the robot at time t is not only related to the received stimulus, but also related to the emotional state at time t-1. Using transformation matrix M1,M2,M3,...,MtObtaining an emotion conversion matrix phi at the moment tt＝M1M2M3…MtThen calculate
Figure BDA00014234138200000315
The emotional state at time t can be obtained. By use of 2m-1 emotion transformation matrix can be usedEach time in the quantum emotion space initializes a unique emotion state, and the emotion space states after all operations are completed are shown in FIG. 3.
Step 4, utilizing quantum computation and projection measurement commonly used in quantum information to carry out entanglement state 2 in quantum emotion spacemThe measurement is carried out for each elementary state. Each measurement can obtain information of a basic state from the quantum emotion space, wherein the information comprises time information and the emotional state of the corresponding moment. When all the information is retrieved, the original quantum emotion space can be obtained.

Claims (1)

1. A method for obtaining a robot emotion space based on a quantum structure is characterized by comprising the following steps: firstly, in a quantum system, constructing a quantum psychological judgment space on the basis of a classical psychological judgment space; the quantum psychology judgment space is a two-dimensional coordinate system, the coordinate axes respectively represent excitement and happiness, and each emotion is quantized into a corresponding value in the coordinate system; the quantum psychology judgment space uses 2n quantum bits to represent position information of the emotion in a coordinate system, wherein the first n quantum bits are used for representing information of the emotion along an excitation coordinate axis, and the last n quantum bits are used for representing information of the emotion along a happy coordinate axis; quantum Emotion Space (QES) is composed of a Quantum psychological judgment space and a time axis, wherein the time axis is expressed by m Quantum bits; using 2n + m quantum bits, the information of QES is integrated into one normalized quantum state, i.e. quantum emotion space expression | e (t) > is defined as:
Figure FDA0002715739170000011
wherein, | t>Representing the time information of each emotion in QES, | At>And | Pt>Respectively represent | t>The position information of the time emotion in the excitation coordinate axis and the happy coordinate axis is defined as:
Figure FDA0002715739170000012
Figure FDA0002715739170000013
wherein the content of the first and second substances,
Figure FDA0002715739170000014
and
Figure FDA0002715739170000015
in a substantially quantum state, i.e.
Figure FDA0002715739170000016
The method comprises the following steps:
step 1, preparing 2n + m quantum bits in quantum system, the state of the quantum system is 2n + m |0>Is the tensor product of
Figure FDA0002715739170000017
Performing Hadamard transform, H transform, on m qubits representing time information, the H transform using unitary matrices, the expression being
Figure FDA0002715739170000018
Separately H-transforming each of the m qubits to generate 2mSuperposition of individual quantum states, i.e. producing from 0 to 2m-1 all binary values, each elementary state existing simultaneously with a probability of being present
Figure FDA0002715739170000019
In the process of performing H-transform considering the integrity of the quantum system, a special two-dimensional identity matrix I is used for processing other 2n quantum bits, wherein the expression of I is
Figure FDA00027157391700000110
By conversion, the time period of the robot emotion space is initialized, namely, the time period is changed
Figure FDA00027157391700000111
Is converted into
Figure FDA00027157391700000112
Step 2, according to the conversion result of step 1, the quantum system at this time comprises 2mEach basic state is composed of 2n quantum bits representing emotional information and unique time information, no emotional state exists in the emotional space of the robot, and when the robot receives an instant stimulus from the outside, the quantum system converts the stimulus into a conversion matrix M identified by the emotional space1Since the fundamental state in the stable quantum system is in the entangled state, the unitary matrix is used to ensure that other 2 s are in the process of initializing the emotionm1 elementary state is not affected; in quantum computing, |1 is utilized><1| Pair of entangled quantum states |1>Projection is obtained
Figure FDA00027157391700000113
Simultaneous calculation
Figure FDA00027157391700000114
That is, obtaining the emotional state at the time t-1, and then using the emotional state
Figure FDA0002715739170000021
Making entangled quantum states respectively pair i>Projecting to obtain the basic states at other moments;
step 3, the emotional state of the robot has continuity and time correlation, and the emotional state of the robot at the time t is not only related to the received stimulus, but also related to the emotional state at the time t-1; using transformation matrix M1,M2,M3,...,MtObtaining an emotion conversion matrix phi at the moment tt，Φt＝M1M2M3...MtThen calculate
Figure FDA0002715739170000022
Obtaining emotional State at time t, use 2m-1 emotion transformation matrix initializing a unique emotion state for each time in the quantum emotion space;
step 4, utilizing quantum computation and projection measurement in quantum information to carry out entanglement state 2 in quantum emotion spacemAnd measuring the basic states, obtaining information of one basic state from the quantum emotion space in each measurement, wherein the information comprises time information and the emotion state at a corresponding moment, and obtaining the original quantum emotion space after all the information is retrieved.
CN201710902831.1A 2017-09-29 2017-09-29 Robot emotion space modeling method based on quantum structure Expired - Fee Related CN107633292B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710902831.1A CN107633292B (en) 2017-09-29 2017-09-29 Robot emotion space modeling method based on quantum structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710902831.1A CN107633292B (en) 2017-09-29 2017-09-29 Robot emotion space modeling method based on quantum structure

Publications (2)

Publication Number Publication Date
CN107633292A CN107633292A (en) 2018-01-26
CN107633292B true CN107633292B (en) 2021-03-09

Family

ID=61102860

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710902831.1A Expired - Fee Related CN107633292B (en) 2017-09-29 2017-09-29 Robot emotion space modeling method based on quantum structure

Country Status (1)

Country Link
CN (1) CN107633292B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108875956B (en) * 2017-05-11 2019-09-10 广州异构智能科技有限公司 Primary tensor processor

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473111A (en) * 2013-08-16 2013-12-25 运软网络科技(上海)有限公司 Brain-imitation calculation virtualization method and brain-imitation calculation virtualization system
US20160189053A1 (en) * 2014-12-31 2016-06-30 Invent.ly LLC Freezing a subject's contextualization of a proposition considered in a quantum representation by using a Quantum Zeno effect
CN104618031B (en) * 2015-02-12 2017-06-09 四川师范大学 The method of the bi-direction controlled quantum teleportation of unknown any two particle
CN106683041B (en) * 2016-12-12 2023-03-07 长春理工大学 Quantum image miscut method based on NEQR expression
CN106956271B (en) * 2017-02-27 2019-11-05 华为技术有限公司 Predict the method and robot of affective state

Also Published As

Publication number Publication date
CN107633292A (en) 2018-01-26

Similar Documents

Publication Publication Date Title
Autthasan et al. MIN2Net: End-to-end multi-task learning for subject-independent motor imagery EEG classification
Biswas et al. Chaos theory and its applications in our real life
CN100415159C (en) Dynamic characteristic analysis method of real-time tendency of heart state
Yilmaz et al. A novel action recognition framework based on deep-learning and genetic algorithms
Kumar et al. MobiHisNet: A lightweight CNN in mobile edge computing for histopathological image classification
Yeomans et al. Simulating time-series data for improved deep neural network performance
CN111028319A (en) Three-dimensional non-photorealistic expression generation method based on facial motion unit
CN106503659A (en) Action identification method based on sparse coding tensor resolution
Tatarian et al. Stepping-stones to transhumanism: An emg-controlled low-cost prosthetic hand for academia
Tong et al. Learn the temporal-spatial feature of sEMG via dual-flow network
CN111709284B (en) Dance emotion recognition method based on CNN-LSTM
CN107633292B (en) Robot emotion space modeling method based on quantum structure
Xu et al. 1.2 watt classification of 3D voxel based point-clouds using a CNN on a neural compute stick
Mousas et al. Learning motion features for example-based finger motion estimation for virtual characters
Lv et al. Cognitive computing for brain–computer interface-based computational social digital twins systems
Chen et al. A novel imbalanced dataset mitigation method and ECG classification model based on combined 1D_CBAM-autoencoder and lightweight CNN model
CN111062621A (en) AIGAN-based action planning method for original observation data
Usman et al. Skeleton-based motion prediction: A survey
Latha et al. A novel and enhanced facial electromyogram based human emotion recognition using convolution neural network model with multirate signal processing features
Nowinski Toward the human nanoscale connectome: Neuronal morphology format, modeling, and storage requirement estimation
He et al. Emotion classification using eeg data in a brain-inspired spiking neural network
JP2021117977A (en) Information processing system, combinatorial optimization computation method, and combinatorial optimization computation program
Dutta et al. An extensive analysis on deep neural architecture for classification of subject-independent cognitive states
Ye et al. Cyclic Convolutional Neural Network Model Based on Artificial Intelligence
JP6401758B2 (en) Tactile content generating device, tactile content generating method, and tactile content using device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20210309