CN108846477B - Intelligent brain decision system and decision method based on reflection arcs - Google Patents
Intelligent brain decision system and decision method based on reflection arcs Download PDFInfo
- Publication number
- CN108846477B CN108846477B CN201810690861.5A CN201810690861A CN108846477B CN 108846477 B CN108846477 B CN 108846477B CN 201810690861 A CN201810690861 A CN 201810690861A CN 108846477 B CN108846477 B CN 108846477B
- Authority
- CN
- China
- Prior art keywords
- decision
- module
- nerve
- arc
- feedback
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/063—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means
- G06N3/065—Analogue means
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Neurology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention relates to an intelligent brain decision system and a decision method based on reflection arcs, wherein the decision system comprises: the reflective arc decision module is used for performing conditional reflection on the acquired decision task and outputting a decision result; the information center is connected with the reflection arc decision module and provides external information for the reflection arc decision module; the feedback module is connected with the reflection arc decision module and used for acquiring feedback information of the reflection arc decision module; the memory module is connected with the reflective arc decision module and records the decision result of the reflective arc decision module; and the model module is respectively connected with the reflective arc decision module, the feedback module and the memory module and is used for carrying out evolution training on the reflective arc decision module based on the feedback information and the decision result. Compared with the prior art, the method has the advantages of wide applicability, capability of solving the problem of decision island and the like.
Description
Technical Field
The invention relates to the field of computer science, in particular to an intelligent brain decision system and a decision method based on reflection arcs.
Background
The development of technologies such as mobile internet, internet of things and AI is pushing mankind to the era of intelligent internet, information shows explosive growth, and big data, a powerful new natural resource, develops rapidly in quantity, diversity and complexity, so that the cost of information screening becomes higher and higher. Predictive analysis technologies such as machine learning and deep learning can improve the capability of data insight, however, in many cases, a decision maker cannot rely on a single decision method to cope with multi-dimensional digital change.
The prior decision system has the following defects: (1) the decision model technology system is complicated, the service forms are different, the decision island phenomenon is obvious, and the comprehensive use of a plurality of models is difficult; (2) decision flows are different and lack of standard constraint, and the repeated work of establishing a decision system is obvious; (3) the concept used in the decision making has higher specialization degree and higher communication sharing difficulty, which is not beneficial to fully exerting the collective advantage; (4) the decision system is incomplete, only the decision is paid attention to, and the evaluation of decision effect, actual feedback tracking and continuous evolution of decision are not considered.
Disclosure of Invention
The present invention is directed to overcoming the above-mentioned drawbacks of the prior art and providing an intelligent brain decision system based on reflex arc.
The purpose of the invention can be realized by the following technical scheme:
a reflex arc-based intelligent brain decision making system, comprising:
the reflective arc decision module is used for performing conditional reflection on the acquired decision task and outputting a decision result;
the information center is connected with the reflection arc decision module and provides external information for the reflection arc decision module;
the feedback module is connected with the reflection arc decision module and used for acquiring feedback information of the reflection arc decision module;
the memory module is connected with the reflective arc decision module and records the decision result of the reflective arc decision module;
and the model module is respectively connected with the reflective arc decision module, the feedback module and the memory module and is used for carrying out evolution training on the reflective arc decision module based on the feedback information and the decision result.
Further, the reflex arc decision module comprises an afferent nerve, a perception nerve, a precursor nerve, a decision nerve, a posterior nerve and an efferent nerve which are connected in sequence, wherein,
the afferent nerves are used for receiving decision tasks;
the perception nerve is connected with the information center and is used for collecting external information;
pre-screening and pre-judging the conditional data of the precursor nerve;
the decision nerve is connected with the model module and used for realizing quantitative decision;
the back-driving nerve is connected with a memory module and is used for judging and intervening the quantitative decision and transmitting the decision result to the memory module;
the efferent nerves are used for outputting decision results.
Further, the reflex arc decision module forms different reflex arcs by combination of different numbers of nerves, each of the trained reflex arcs representing a decision strategy.
Further, the reflective arc structure includes a simple reflective structure and a complex reflective structure.
Further, each of the reflex arcs includes only one afferent nerve, one efferent nerve, one memory nerve, one feedback nerve, and one evolutionary nerve.
Further, the feedback module comprises a feedback nerve and a feedback center connected, wherein,
the feedback nerve is connected with the reflection arc decision module and used for collecting feedback information;
the feedback center is connected with the model module, and is used for uniformly managing the feedback information and transmitting the feedback information to the model module.
Further, the memory module comprises a memory nerve and a memory center, wherein,
the memory nerve is connected with the reflex arc decision module and used for recording a decision result;
and the memory center is connected with the model module, and is used for uniformly managing the decision result and transmitting the decision result to the model module.
Further, the model module comprises a connected evolved nerve and a model center, wherein,
the evolutionary nerves are respectively connected with a feedback module and a memory module, and the evolutionary training is carried out on the reflex arc decision module based on the feedback information and the decision result;
the center of the model is connected with the reflective arc decision module, and the trained reflective arc model is managed in a unified mode.
The invention also provides a decision-making method of the intelligent brain decision-making system based on the reflex arcs, which carries out a conditional reflex response decision-making task by the trained reflex arc decision-making module, records feedback information and decision-making results in the process of conditional reflex, and carries out optimization training on the reflex arc decision-making module.
Compared with the prior art, the invention has the following beneficial effects:
1) the invention realizes different reflecting arcs by utilizing the flexible combination of the afferent nerves, the perception nerves, the prodromal nerves, the decision nerves, the posterior drive nerves and the efferent nerves, thereby realizing a universal decision scheme and having wide applicability.
2) The invention is a decision system based on the reflex arc and the nervous system, and can more vividly and exactly describe the intelligent brain.
3) The invention adopts a scheme of model integration and comprehensive decision, and provides an intelligent scheme for enhancing the intelligent brain.
4) The method has the advantages that the decision making process is realized, the memory and feedback information are integrated, the scattered knowledge segments are connected, the data quality is effectively improved, the continuous optimization progress of a decision making model is realized, a more sufficient training sample is further provided, and data and algorithm support is provided for the continuous evolution of the reflection arc.
5) The invention provides the ideas of unifying data, unifying algorithms, unifying decisions and unifying evolution, breaks through the data, algorithms and decision islands, has less repeated work and improves the resource utilization rate.
Drawings
FIG. 1 is a schematic view of the structure of the present invention;
FIG. 2 is a diagram of an example of a champion challenger mode of the present invention;
FIG. 3 is a diagram of an embodiment of a multi-model integrated decision making method according to the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
As shown in fig. 1, the present invention provides a reflective arc-based intelligent brain decision system, which comprises a reflective arc decision module, an information center, a feedback module, a memory module and a model module, wherein the reflective arc decision module performs conditional reflection on an acquired decision task and outputs a decision result; the information center 7 is connected with the reflection arc decision module and provides external information for the reflection arc decision module; the feedback module is connected with the reflection arc decision module to acquire feedback information of the reflection arc decision module; the memory module is connected with the reflective arc decision module and records the decision result of the reflective arc decision module; the model module is respectively connected with the reflection arc decision module, the feedback module and the memory module, and evolution training is carried out on the reflection arc decision module based on feedback information and decision results.
The reflex arc decision module comprises an afferent nerve 1, a perception nerve 2, a precursor nerve 3, a decision nerve 4, a posterior drive nerve 5 and an efferent nerve 6 which are sequentially connected, wherein the afferent nerve 1 is used for receiving a decision task; the perception nerve 2 is connected with the information center 7 and is used for collecting external information; the precursor nerve 3 realizes the pre-screening and pre-judging of condition data; the decision nerve 4 is connected with the model module and is used for realizing quantitative decision; the back drive nerve 5 is connected with the memory module and is used for judging and intervening the quantitative decision and transmitting the decision result to the memory module; the efferent nerve 6 is used to output a decision result.
The reflex arc decision module forms different reflex arcs by combining different numbers of nerves, each of the trained reflex arcs represents a decision strategy, and the combination of the reflex arcs obeys constraint rules known in the art. The reflection arc structure comprises a simple reflection structure and a complex reflection structure, the reflection made by the reflection arc composed of the single model service is a simple reflection, and the rest are complex reflections. Each reflex arc comprises only one afferent nerve, one efferent nerve, one memory nerve, one feedback nerve and one evolutionary nerve, and there can be a plurality of perception nerves, precursor nerves, decision nerves and posterior nerves.
The feedback module comprises a feedback nerve 8 and a feedback center 9 which are connected, wherein the feedback nerve 8 is connected with the reflex arc decision module and is used for collecting feedback information; the feedback center 9 is connected with the model module, and is used for uniformly managing the feedback information and transmitting the feedback information to the model module.
The memory module comprises a memory nerve 10 and a memory center 11, wherein the memory nerve 10 is connected with the reflective arc decision module and used for recording a decision result; the memory center 11 is connected with the model module, and manages the decision result in a unified way and transmits the decision result to the model module.
The model module comprises an evolutionary nerve 12 and a model center 13 which are connected, wherein the evolutionary nerve 12 is respectively connected with the feedback module and the memory module, and the evolutionary training is carried out on the reflex arc decision module based on the feedback information and the decision result; the model center 13 is connected with the reflective arc decision module to perform unified management on the trained reflective arc model.
The working principle of the decision system comprises the following steps:
(1) and (5) establishing a decision strategy. Afferent nerves, perception nerves, precursor nerves, decision nerves, rear-driving nerves, memory nerves, efferent nerves, feedback nerves and evolutionary nerves flexibly combine into reflection arcs with different complexity degrees under the condition of complying with the constraint rules so as to respond to the problem reflection with different complexity degrees, and the interrelation is shown in table 1.
TABLE 1
(2) And (5) a decision response process. Each stimulus to the intelligent brain will respond by a reflex of the corresponding reflex arc. The reflective arcs can only make conditional reflections, i.e. the reflections they make are all reflective arcs formed by training.
(3) And (5) deciding an evolution process. The evolutionary nerves integrate memory and feedback information, scattered knowledge segments are connected, data quality is improved, a more sufficient training sample is provided for continuous optimization progress of an algorithm model, and data and algorithm support are provided for continuous evolution of a reflex arc.
Examples
(1) The implementation of the decision strategy of the champion challenger mode in the decision system is illustrated by taking classified response to the user as an example. A reflex arc decision module of the champion challenger pattern as shown in fig. 2 was first established with 1 sensory nerve, 1 prodromal nerve, 3 decision nerves and 1 postero-repellent nerve. Afferent nerves obtain a certificate number; the perception nerve acquires user information according to the certificate number; the precursor nerve prejudges user data, configures rules of a champion challenger mode, and determines a decision model to which each request should be distributed; the 3 decision nerves are respectively a champion model and a challenger model, the same function is realized, decision requests with different proportions are respectively processed, if the effect of the challenger model is better than that of the champion model, the challenger model can replace the champion model, and the optimal model selection is realized by continuous evolution; the memory nerve and the feedback nerve respectively record a decision result and actual feedback of a user, and a more sufficient training sample is provided for the evolution of the evolutionary nerve.
(2) The implementation of the multi-model comprehensive decision strategy in the decision system is illustrated by taking the decision response to the certificate number as an example. Firstly, establishing a multi-model mode reflection arc decision module as shown in fig. 3, wherein the module is provided with 3 perception nerves, 2 precursor nerves, 3 decision nerves and 2 back-drive nerves, the afferent nerves acquire certificate numbers, the perception nerves a and b acquire company names and department position information according to the certificate numbers, and the decision nerves a and b respectively predict industry codes and occupational codes; the perception nerve c obtains consumption rating information according to the certificate number, and the decision nerve c predicts a risk label according to the consumption rating information, the industry code and the occupation code. The memory nerve and the feedback nerve respectively record a decision result and actual feedback of a user, and a more sufficient training sample is provided for the evolution of the evolutionary nerve.
The foregoing detailed description of the preferred embodiments of the invention has been presented. It should be understood that numerous modifications and variations could be devised by those skilled in the art in light of the present teachings without departing from the inventive concepts. Therefore, the technical solutions that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the prior art according to the concepts of the present invention should be within the scope of protection determined by the claims.
Claims (7)
1. An intelligent brain decision system based on reflex arcs, comprising:
the reflective arc decision module is used for performing conditional reflection on the acquired decision task and outputting a decision result;
the information center is connected with the reflection arc decision module and provides external information for the reflection arc decision module;
the feedback module is connected with the reflection arc decision module and used for acquiring feedback information of the reflection arc decision module;
the memory module is connected with the reflective arc decision module and records the decision result of the reflective arc decision module;
the model module is respectively connected with the reflective arc decision module, the feedback module and the memory module and is used for carrying out evolution training on the reflective arc decision module based on the feedback information and the decision result;
the reflex arc decision module comprises an afferent nerve, a perception nerve, a precursor nerve, a decision nerve, a posterior drive nerve and an efferent nerve which are connected in sequence,
the afferent nerves are used for receiving decision tasks;
the perception nerve is connected with the information center and used for collecting external information;
pre-screening and pre-judging the precursor nerve realization condition data;
the decision nerve is connected with the model module and used for realizing quantitative decision;
the back drive nerve is connected with the memory module and is used for judging and intervening the quantitative decision and transmitting the decision result to the memory module;
the efferent nerve is used for outputting a decision result;
the feedback module comprises a feedback nerve and a feedback center which are connected, wherein,
the feedback nerve is connected with the reflection arc decision module and used for collecting feedback information;
the feedback center is connected with the model module, and is used for uniformly managing the feedback information and transmitting the feedback information to the model module;
when the system is used for realizing risk label prediction, a nerve is introduced to obtain a certificate number, a perception nerve obtains consumption rating information, a company name and department position information according to the certificate number, a decision nerve predicts an industry code and a occupation code respectively according to the company name and the department position information, a risk label is predicted according to the consumption rating information, the industry code and the occupation code, and a memory nerve and a feedback nerve record a decision result and actual feedback of a user respectively.
2. The intelligent reflex-arc-based brain decision system according to claim 1, wherein the reflex arc decision module forms different reflex arcs by combination of different numbers of nerves, each of the trained reflex arcs representing a decision strategy.
3. The reflective arc based intelligent brain decision system according to claim 2, wherein the reflective arc structures comprise simple reflective structures and complex reflective structures.
4. The intelligent reflex-based brain decision making system as claimed in claim 2, wherein each reflex arc includes only one afferent nerve, one efferent nerve, one memory nerve, one feedback nerve and one evolutionary nerve.
5. The intelligent reflex arc-based brain decision system according to claim 1, wherein the memory module includes a memory nerve and a memory center, wherein,
the memory nerve is connected with the reflecting arc decision module and used for recording a decision result;
and the memory center is connected with the model module, and is used for uniformly managing the decision result and transmitting the decision result to the model module.
6. The intelligent reflex arc-based brain decision system of claim 1, wherein the model module comprises a connected evolutionary nerve and a model center, wherein,
the evolutionary nerves are respectively connected with a feedback module and a memory module, and the evolutionary training is carried out on the reflex arc decision module based on the feedback information and the decision result;
the model center is connected with the reflective arc decision module to carry out unified management on the trained reflective arc model.
7. A decision making method of the reflex arc-based intelligent brain decision making system according to claim 1, wherein the method performs a conditioned reflex response decision making task by a trained reflex arc decision making module, records feedback information and decision making results during the conditioned reflex process, and performs optimization training on the reflex arc decision making module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810690861.5A CN108846477B (en) | 2018-06-28 | 2018-06-28 | Intelligent brain decision system and decision method based on reflection arcs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810690861.5A CN108846477B (en) | 2018-06-28 | 2018-06-28 | Intelligent brain decision system and decision method based on reflection arcs |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108846477A CN108846477A (en) | 2018-11-20 |
| CN108846477B true CN108846477B (en) | 2022-06-21 |
Family
ID=64201010
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810690861.5A Active CN108846477B (en) | 2018-06-28 | 2018-06-28 | Intelligent brain decision system and decision method based on reflection arcs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108846477B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110941797B (en) * | 2019-11-07 | 2023-04-07 | 中信银行股份有限公司 | Operation index monitoring and trend prediction system based on service index |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4818234A (en) * | 1986-06-25 | 1989-04-04 | Redington Dana J | Psychophysiological reflex arc training simulator |
| CN101599137A (en) * | 2009-07-15 | 2009-12-09 | 北京工业大学 | Autonomous operant conditioning reflex automat and the application in realizing intelligent behavior |
| CN103026973A (en) * | 2012-11-30 | 2013-04-10 | 中国航天员科研训练中心 | Device for detecting, analyzing and processing reward operation condition reflection in real time |
| CN105279555A (en) * | 2015-10-28 | 2016-01-27 | 清华大学 | Self-adaptive learning neural network implementation method based on evolutionary algorithm |
| CN105427014A (en) * | 2015-10-27 | 2016-03-23 | 余华鸿 | Method and system used for emergency decision of nuclear power accidents |
| CN106560848A (en) * | 2016-10-09 | 2017-04-12 | 辽宁工程技术大学 | Novel neural network model for simulating biological bidirectional cognition capability, and training method |
| CN107844869A (en) * | 2017-11-30 | 2018-03-27 | 中国华电集团科学技术研究总院有限公司 | The on-line intelligence learning decision optimization method and system of gas distributed energy system |
-
2018
- 2018-06-28 CN CN201810690861.5A patent/CN108846477B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4818234A (en) * | 1986-06-25 | 1989-04-04 | Redington Dana J | Psychophysiological reflex arc training simulator |
| CN101599137A (en) * | 2009-07-15 | 2009-12-09 | 北京工业大学 | Autonomous operant conditioning reflex automat and the application in realizing intelligent behavior |
| CN103026973A (en) * | 2012-11-30 | 2013-04-10 | 中国航天员科研训练中心 | Device for detecting, analyzing and processing reward operation condition reflection in real time |
| CN105427014A (en) * | 2015-10-27 | 2016-03-23 | 余华鸿 | Method and system used for emergency decision of nuclear power accidents |
| CN105279555A (en) * | 2015-10-28 | 2016-01-27 | 清华大学 | Self-adaptive learning neural network implementation method based on evolutionary algorithm |
| CN106560848A (en) * | 2016-10-09 | 2017-04-12 | 辽宁工程技术大学 | Novel neural network model for simulating biological bidirectional cognition capability, and training method |
| CN107844869A (en) * | 2017-11-30 | 2018-03-27 | 中国华电集团科学技术研究总院有限公司 | The on-line intelligence learning decision optimization method and system of gas distributed energy system |
Non-Patent Citations (2)
| Title |
|---|
| "基于决策神经科学的风险决策机理研究";沈强;《中国博士学位论文全文数据库 经济与管理科学辑》;20111015;摘要 * |
| "基于智能策略的网络自管理模型及应用研究";刘素平;《中国博士学位论文全文数据库 信息科技辑》;20091015;第49-55、84页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108846477A (en) | 2018-11-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Vakili et al. | Comprehensive and systematic review of the service composition mechanisms in the cloud environments | |
| Berger et al. | The Nature of Digital Technologies-Development of a Multi-Layer Taxonomy. | |
| AU2018201691B2 (en) | Job allocation | |
| Zhang et al. | Toward Green Metaverse Networking: Technologies, Advancements, and Future Directions | |
| CN109992404A (en) | PC cluster resource regulating method, device, equipment and medium | |
| CN110147389A (en) | Account number treating method and apparatus, storage medium and electronic device | |
| CN110610346A (en) | Intelligent office automation system workflow instance time prediction analysis | |
| CN108846477B (en) | Intelligent brain decision system and decision method based on reflection arcs | |
| CN111582341A (en) | User abnormal operation prediction method and device | |
| CN113222560A (en) | Enterprise service system based on AI consultation and construction method thereof | |
| Chanthati | Artificial Intelligence-Based Cloud Planning and Migration to Cut the Cost of Cloud Sasibhushan Rao Chanthati | |
| CN113138847A (en) | Computer resource allocation scheduling method and device based on federal learning | |
| Lin | The theoretical framework of enterprise digital innovation: insights from a qualitative meta-analysis | |
| Priore et al. | Learning-based scheduling of flexible manufacturing systems using support vector machines | |
| Shukla et al. | Analysis of critical drivers affecting implementation of agent technology in a manufacturing system | |
| Khan et al. | Dynamic offloading technique for real-time edge-to-cloud computing in heterogeneous MEC–MCC and IoT devices | |
| US20180357586A1 (en) | Systems and methods for holistically and dynamically providing quintessential conseiller functionality | |
| Zhong et al. | Analysis of enterprise strategic management issues and coping strategies based on big data analysis | |
| Monghasemi et al. | A hybrid of genetic algorithm and evidential reasoning for optimal design of project scheduling: A systematic negotiation framework for multiple decision-makers | |
| Kavitha et al. | Ai Integration in Data Driven Decision Making for Resource Management in Internet of Things (Iot): A Survey | |
| Appadurai et al. | Radial basis function networks-based resource-aware offloading video analytics in mobile edge computing | |
| Zhu et al. | A Resource scheduling method for enterprise management based on artificial intelligence deep learning | |
| Xiao et al. | Mobile-edge-platooning cloud: a lightweight cloud in vehicular networks | |
| Tardivo et al. | Hierarchical parallel model for improving performance on differential evolution | |
| CN115827232A (en) | Method, device, system and equipment for determining configuration for service model |
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 |