Current Issue

15 June 2023, Volume 5 Issue 2

Previous Issue    Next Issue

Review Intelligence

In celebration of McCulloch-Pitts ANN model’s 80th anniversary: its origin, principle, and influence

Qinghai MIAO, Yutong WANG, Yisheng LV, Xiaoxiang NA, Fei-Yue WANG

2023, 5(2):  133-142.  doi:10.11959/j.issn.2096-6652.202314

Asbtract ( 539 )   HTML ( 140)   PDF (8731KB) ( 501 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In 1943, Warren McCulloch and Walter Pitts published a paper titled “A logical calculus of the ideas immanent in nervous activity”, which demonstrated that the functioning of neural networks could be described using logical calculus.This expanded the field of computational neuroscience and laid the foundation for the development of artificial neural networks.On the occasion of the 80th anniversary of the publication of the M-P paper, the intellectual origins of the M-P theory was explored, taking into account the historical context and the authors’ career trajectories.With the help of examples from the original paper, the basic principles of the M-P model were outlined, and its advantages and limitations were summarized.Furthermore, the impact of the M-P theory on the development of information science was discussed, focusing on the authors’ contributions to cybernetics, including their work on circular causality and feedback mechanisms, which provided a foundation for the development of modern artificial intelligence technologies such as parallel intelligence and large-scale models.

Surveys and Prospectives

Key problems and progress of pedestrian trajectory prediction methods: the state of the art and prospects

Quancheng DU, Xiao WANG, Lingxi LI, Huansheng NING

2023, 5(2):  143-162.  doi:10.11959/j.issn.2096-6652.202315

Asbtract ( 528 )   HTML ( 88)   PDF (5939KB) ( 471 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Pedestrian trajectory prediction aims to use observed human historical trajectories and surrounding environmental information to predict the future position of the target pedestrian, which has important application value in reducing collision risks for autonomous vehicles in social interactions.However, traditional model-driven pedestrian trajectory prediction methods are difficult to predict pedestrian trajectories in complex and highly dynamic scenes.In contrast, datadriven pedestrian trajectory prediction methods rely on large-scale datasets and can better capture and model more complex pedestrian interaction relationships, thereby achieving more accurate pedestrian trajectory prediction results, and have become a research hotspot in fields such as autonomous driving, robot navigation and video surveillance.In order to macroscopically grasp the research status and key issues of pedestrian trajectory prediction methods, We started with the classification of pedestrian trajectory prediction technology and methods.First, the research progress of existing pedestrian trajectory prediction methods were elaborated and the current key issues and challenges were summarized.Second, according to the modeling differences of pedestrian trajectory prediction models, existing methods were divided into model-driven and data-driven pedestrian trajectory prediction methods, and the advantages, disadvantages and applicable scenarios of different methods were summarized.Then, the mainstream datasets used in pedestrian trajectory prediction tasks were summarized and the performance indicators of different algoriths were compared.Finally, the future development direction of pedestrian trajectory prediction was prospected.

A survey on network tomography technology

Xiaojia XU, Yongcai WANG, Deying LI

2023, 5(2):  163-179.  doi:10.11959/j.issn.2096-6652.202316

Asbtract ( 133 )   HTML ( 26)   PDF (4933KB) ( 265 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Network tomography is an efficient and convenient tool to infer the internal state of a network through end-toend path measurement.It has been widely used in link measurement and fault location of wired and wireless networks.This paper summarized and analyzed the model foundation of four basic models of network tomography, boolean network tomography technology, additive network tomography technology, bandwidth network tomography technology and the stochastic network tomography technology.This paper systematically combed the key problems in network tomography technology, including monitor placement, beacon and service placement, path construction and data analysis, the problem of identifiability, and attack in network tomography when identification was not guaranteed.At the same time, the new research and new problems in the field of network tomography such as network tomography with network coding, neural network tomography and node fault location method in NFV were sorted out.Finally, based on the development status of network tomography technology, the future development trend of network tomography was analyzed and discussed.

Deep learning-based multimodal trajectory prediction methods for autonomous driving: state of the art and perspectives

Jun HUANG, Yonglin TIAN, Xingyuan DAI, Xiao WANG, Zhixing PING

2023, 5(2):  180-199.  doi:10.11959/j.issn.2096-6652.202317

Asbtract ( 865 )   HTML ( 102)   PDF (6643KB) ( 619 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Although deep learning methods have achieved better results than traditional trajectory prediction algorithms, there are still problems such as information loss, interaction and uncertainty difficulties in modelling, and lack of interpretability of predictions when implementing multimodal high-precision prediction for autonomous vehicles in heterogeneous, highly dynamic and complex changing environments.The newly developed Transformer's long-range modelling capability and parallel computing ability make it a great success not only in the field of natural language processing, but also in solving the above problems when extended to the task of multimodal trajectory prediction for autonomous driving.Based on this, the aim of this paper is to provide a comprehensive summary and review of past deep neural network-based approaches, in particular the Transformer-based approach.The advantages of Transformer over traditional sequential network, graphical neural network and generative model were also analyzed and classified in relation to existing challenges, simultaneously.Transformer models can be better applied to multimodal trajectory prediction tasks, and that such models have better generalisation and interpretability.Finally, the future directions of multimodal trajectory prediction were presented.

A survey on application of block chain in next generation intelligent manufacturing

Zhiwei LIN, Songchuan ZHANG, Chengji WANG, Yiwei ZHOU

2023, 5(2):  200-211.  doi:10.11959/j.issn.2096-6652.202248

Asbtract ( 460 )   HTML ( 65)   PDF (13396KB) ( 477 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In the process of promoting digital transformation of Chinese manufacturing enterprises, block chain technology has attracted extensive attention of researchers.The overall research progress domestically and abroad in the field of block chain technology was concluded.The recent researches on block chain technology in manufacturing process of food and military products, cloud manufacturing and other fields were summarized.The future path of combining block chain technology with intelligent manufacturing was predicted from 4 aspects: storage problem, scalability problem, computational power waste problem, security and privacy problem, aiming to provide valuable research ideas for further researchers.

Special Topic: Frontier of R＆D in Parallel Intelligence

Parallel scientific research institutes: from digital transformation to intelligent revolution

Rui QIN, Xiaolong LIANG, Juanjuan LI, Wenwen DING, Jiachen HOU, Yutong WANG, Yonglin TIAN, Ding WEN

2023, 5(2):  212-221.  doi:10.11959/j.issn.2096-6652.202318

Asbtract ( 225 )   HTML ( 27)   PDF (8131KB) ( 387 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Aiming to address the dual complexities that current scientific research institutes face in both management work and research tasks, parallel scientific research institutes were proposed.Parallel scientific research institutes were constructed based on the virtual-real interactive parallel intelligence theory, which leveraged digital construction technologies based on digital twins and metaverse, distributed governance technologies based on blockchain and decentralized autonomous organizations and operations (DAOs), intelligent decision-making technologies based on big data and foundation models, as well as scientific innovation paradigms based on decentralized science (DeSci) and artificial intelligence for science (AI4S).Their core goal was to form a prescriptive scheme for scientific research institute transformation driven by complex science, thereby constructing a trustworthy, reliable, usable, efficient and effective intelligent research organizational and operational ecosystem.The systematic design and key technologies of parallel scientific research institutes were introduced, their main characteristics and advantages were described, and their typical application scenarios were explored.Parallel scientific research institutes go beyond the simple digital transformation of scientific research institutes, and emphasize a higher level of intelligent transformation, promoting the sustainable and healthy development of scientific research institutes.

Parallel surgical robots

Tianxiang BAI, Shuangyi WANG, Yating LIU, Hanzhong LI, Yi WEN

2023, 5(2):  222-233.  doi:10.11959/j.issn.2096-6652.202319

Asbtract ( 246 )   HTML ( 17)   PDF (15489KB) ( 220 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Medical robots improve medical efficiency and reduce patients’ suffering that owns social significance and scientific research value.Focusing on the difficulties of medical robots, we propose the concept of parallel medical robotics based on the ACP (artificial societies, computational experiments, parallel execution)-based parallel system method.parallel medical robot is mainly composed of a physical medical robot and its virtual counterpart, through which we can realize the management and control for the overall system, conduct experiments and evaluation for the treatment process, and learning and training for doctors and patients.Parallel medical robotics is supported by robot simulation, biomechanism, 3D printing and knowledge automation techniques, and holds three core functions that are descrition, prediction and prescription when combined with parallel learning.We demonstrate the framework and its potential usages with a robotic trans-esophageal ultrasound system.

Parallel anesthesia: from anesthesia automation to intelligent full-cycle anesthesia platform

Yifei ZHAO, Le SHEN, Peijun YE, Jing WANG, Fei-Yue WANG

2023, 5(2):  234-246.  doi:10.11959/j.issn.2096-6652.202324

Asbtract ( 141 )   HTML ( 13)   PDF (7432KB) ( 172 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Based on the parallel medical theory, this paper proposes the theory and method of parallel anesthesia, which aims to build an organic and unified intelligent full-cycle anesthesia platform composed of biological humans (medical personnel), robots (mechanical automation equipment) and digital humans (digital medical care) through artificial system modeling, computational experimental analysis and parallel execution of human-computer interaction.Starting from the current status quo and bottleneck of the development of anesthesiology, this paper first conceives multiple intelligent scene modes jointly constructed by biological humans, digital humans and robots in parallel anesthesia systems from the aspects of clinical anesthesia management, crisis management, scientific research, education and teaching, and operation management, and analyzes the role of biological humans in each scene and the complementary relationship between the three.At the same time, the improvement and improvement of parallel anesthesia on patients’ medical safety and medical work efficiency through computational experiments and the implementation correction mode of virtual and real combination are introduced.Finally, by analyzing the ethical issues in the construction of parallel anesthesia system, analyzing the boundary and regulations of automation and intelligent equipment participation in anesthesia, and taking “respect for life, fairness and transparency, efficient operation and labor saving” as the system criteria, an intelligent full-cycle anesthesia platform is established to reduce the repetitive work of medical mechanization, enhance the ability of intelligent decisionmaking assistance, and improve the level of refined treatment and management.

ParaDefender: the parallel supervision for cybersecurity based on ACP

Xiaoguang CHEN, Jinpeng HAN, Manzhi YANG, Xiao WANG, Xin LIU, Zhen WANG, Fei-Yue WANG

2023, 5(2):  247-253.  doi:10.11959/j.issn.2096-6652.202320

Asbtract ( 142 )   HTML ( 11)   PDF (9044KB) ( 152 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Cybersecurity originated from the theoretical testing of computer networks.With the development of information and communication technology (ICT), cybersecurity threats present the ＆quot;4V＆quot; characteristics of high source volume, many varieties, process not visibility, and high-value losses.This is accompanied by a double change in the formation and complexity of cyberspace.With the recent rising of metaverse, the human-centered cyberspace has the ability to map the real world and exposes the real world to cybersecurity threats.However, cybersecurity supervision does not keep up with the changes in cyberspace and still seeks solutions in cyberspace.A parallel supervision system for cybersecurity, ParaDefender was introduced, which was constructed based on artificial systems, computational experiments and parallel execution of ACP method to improve the current dilemma of cybersecurity supervision from the perspective of cyber-physicalsocial systems (CPSS).We demonstrated provided the application solutions of ParaDefender in telecom anti-fraud and industrial IoT scenarios and inspiration and reference for cybersecurity supervision in different fields.

Parallel workshop scheduling model and system design

Shaoming PENG, Gang XIONG, Zhen SHEN, Xisong DONG, Zhiping QU, Long FU, Zhikun TAO, Yunjun HAN

2023, 5(2):  254-266.  doi:10.11959/j.issn.2096-6652.202321

Asbtract ( 206 )   HTML ( 20)   PDF (6948KB) ( 199 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The workshop environment is a highly dynamic and tightly coupled complex system.A scheduling algorithm cannot be trained once and used for life.It needs to perform computational experiments and progressive learning according to the workshop environment and processing state.In this paper, a parallel workshop scheduling system model was proposed to realize scheduling optimization and system evolution.For this model, we described the modeling of artificial workshop scheduling system based on multi-agent system, the computational experiment for workshop scheduling tasks, and the parallel scheduling method for the virtual-real system, realizing the closed-loop control and iterative optimization.Finally, under the guidance of the parallel workshop scheduling model, we designed the architecture and basic functions of the parallel workshop scheduling system.

Parallel fuzzy control: a self-learning control method with virtual-real interaction and mutual enhancement

Dewang CHEN, Jixiang OU

2023, 5(2):  267-273.  doi:10.11959/j.issn.2096-6652.202322

Asbtract ( 179 )   HTML ( 18)   PDF (1670KB) ( 263 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Fuzzy control has advantages such as interpretability and ease of implementation.However, it is limited by its weak self-learning capability, which makes it difficult to effectively utilize the large amount of data accumulated in the control process.Parallel control is a new intelligent control method that enables intelligent control with virtual-real interaction and mutual enhancement, effectively using the Internet and big data to achieve intelligent control.Fuzzy control and parallel control were combined as a new method, and the definition and framework of parallel fuzzy control were proposed and its possible applications were discussed.Parallel fuzzy control has the potential to extend the development direction of fuzzy control and become a new thinking for parallel control.It can effectively utilize big data and some machine learning algorithms based on data-driven to achieve self-learning control, while ensuring interpretability and credibility.

Special Topic: Frontier of R&D in Parallel Intelligence

Parallel design: non-standard mechanical scheme design procedure in parallel manufacturing

Shimeng LI, Fei-Yue WANG

2023, 5(2):  274-282.  doi:10.11959/j.issn.2096-6652.202323

Asbtract ( 119 )   HTML ( 12)   PDF (2822KB) ( 120 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Parallel manufacturing is a new manufacturing form in industry, deeply integrating informalization, automation, and AI. In this paper we illustrated the procedure of non-standard mechanical design in parallel manufacturing, claiming it a nested parallel system. We proposed the ACP method based on the standard procedure and the emulated procedure, and defined the social-value-vector and trended-social-value-matrix so as to take the advantage of data in CPSS. We also presented Xinglun robot speed reducer as a case study of non-standard mechanical design in parallel design.