Methodology For Complex Systems Research Articles

The Increase of Complexity and its Reduction

Taking the lead from complexity theory and complex systems methodology, the article argues that we are engaged in a contradictory process when encountering, analysing and dealing with complexity. We face opposite tendencies that indicate an in-built dynamic between the increase of complexity and its reduction. The increase partly comes through evolution, defined as the transmission of information and partly from the desire for a human-built world that functions more efficiently. The reduction of complexity is due partly to the necessity of minimizing unwanted and unintended consequences of its increase, and partly due to the continued re-alignment between social systems and their environments. The article examines the public debate about human stem cell research, the debates in the 18th century about the free circulation of commodities and opinions, and the attempts to provide answers to the question debated between Einstein and Freud: why war? The article concludes by arguing that the plea for a re-alignment of scientific disciplines will not suffice. Instead, the study of the co-evolution between science and society offers itself as a strategic research site by focusing on the dynamic interplay between the increase of complexity and its reduction.

Multi-scale methodology for complex systems

This paper reviews the development of the multi-scale methodology in complexity science, and elucidates the importance of understanding complex systems in chemical engineering by discussing the relevance between complex systems and a few challenging problems in this field. Multi-scale structure is considered to be the focus in studying complex systems, particularly, the following four critical issues: correlation between phenomena at different scales, compromise between different dominant mechanisms, coupling between spatial and temporal structural changes and critical phenomena occurring in complex systems. It is considered that complex systems can be formulated as a multi-objective variational problem. The multi-scale methodology is promising in coping with complex systems, which can be descriptive, correlative or variational. It will be reviewed in particular by analyzing the above four critical issues and by showing its 20-year development at IPE from a rough idea to modeling approaches, softwares and finally to industrial applications as well as its extension to a number of chemical and physical systems. The strategy of “ from the particular to the general” in developing the variational multi-scale methodology is emphasized, and challenges for mathematicians and physicists are identified to show the necessity of transdisciplinary cooperations. The paper will be concluded by prospects and suggestions.

ACTIVATION HEBBIAN LEARNING RULE FOR FUZZY COGNITIVE MAPS

Fuzzy Cognitive Maps (FCM) is a soft computing, modeling methodology for complex systems, which is originated from the combination of fuzzy logic and neural networks. Many different learning algorithms have been suggested for the training of neural networks. Only some initial thoughts on learning rules have described for FCMs. A learning law is a mathematical algorithm, which can train the FCM by selecting the appropriate weights and it is very important for a system to have learning and adaptive capabilities. In this paper a new learning algorithm, the Activation Hebbian Learning (AHL) has been proposed for FCMs. The learning rule for a FCM is a procedure where FCM weight matrix is modified in order the FCM to model the behavior of a system. Simulation results proving the strength of the learning rule are provided.

Fuzzy modelling and tracking control of nonlinear systems

This paper presents a fuzzy modelling and tracking control methodology for complex systems by combining the merits of fuzzy logic and conventional linear control theory. Here, fuzzy logic is used to formulate a system model by aggregating a set of linearized local subsystems which identify the nonlinear system approximately, and a fuzzy feedback controller is designed by use of conventional linear feedback theory and fuzzy reasoning. A simulation example of a one-link rigid robotic manipulator is given to demonstrate the validity of the proposed control scheme. It is shown that the fuzzy model can be simplified, and good tracking control performance can be achieved by choosing appropriate fuzzy roles.

Scientific methodology for complex systems: Macroscopic patterns in eco‐ and anthropo‐sphere

AbstractA quite unconventional, innovative scientific methodology called “macroscopic pattern analysis” is presented in this paper. This approach is more adequate in the case of complex systems than the well‐known microscopic, mechanistic approach. Complex systems are not only attracting more engineering interest, but their scientific treatment is increasingly wanted by society due to the manifold problems in Earth's ecosphere. The macroscopic pattern approach will be explained in depth and illustrated in some case studies from the ecosphere (sustainability, hurricanes and avalanches), where nature serves as a teacher for the solution of the sustainability problem. Then, a series of case studies on macropatterns are described showing the problem‐solving capacity for anthropo‐ and technosphere: sustainability in society with an index of sustainability, the eco‐social market economy with eco‐tech as an instrument, biokinetics, bioreactor mixing and integrated bioprocessing with models, design of cars and houses and even quality of life as an attempt to quantify macropatterns.The innovations are briefly compared in their problem‐solving capacity with known approaches such as the microscopic method in science, technology and society (free market economy), including the evaluation of other indices and cleaner production, industrial ecology and zero emission initiative. Finally, a deeper integration of sciences, ethics, arts and nature will be introduced based on the vision with macroscopic pattern analysis, where the different domains of human life are integratable to effect a reconciliation.

“Macroscopic pattern analysis” based on formal analogies as a scientific methodology for complex systems

AbstractBioprocess engineering at present concentrates on the enormous problems in the environment. What is therefore needed is a sound methodology, which should be based on the interactions between the physiology of biological reaction networks and physical processes in the environment and which should be an analogy to bioreactor performance. The conventional methodology is empirically oriented, using pilot plant data for the experimental estimates of process economics, where all further details are elucidated following the mechanistic approach on the microscopic level based on assumed mechanisms (causalities). According to the new view, the new systems‐based methodology uses mathematical models as approximations and includes all the interactions. Pilot plant data are needed for model falsification, using analogies on the formal macroscopic level. Bioreactor scale‐up as one application is a more rapid procedure of reasonable accuracy, where both the biokinetics as well as the fluid dynamics are quantified using formal macroscopic analogies.Model consistency and plausibility are the basic criteria when using model computer simulations as a decisive aid, while experiments lose their central role and are on longer the basis for evaluating the scientific work; they are simply the basis of the researcher's intuition. Another typical feature of complex systems is that model parameters are interdependent. The final decisive fact will be the mental experiment (“thinking” with the left and right side of the brain), which can be supported using computer simulations. This evolutionary interplay between the three realities of thinking, experimenting and simulating leads to a holistic progress towards the better understanding of highly complex systems. It offers the solution to the problems, which is needed in future.

Policy and decision support in sustainable development planning: a ‘complex systems methodology’ for use in sub-Saharan Africa

With continuing emphasis on environmental impact assessment, there is a need to develop more complex models to allow decision- makers to explore more widely the implications of change within human systems. SENSIM is a computer- based interactive geographical information system developed for use in Senegal: it is currently being updated for other areas. It has been designed for decision-makers themselves to use, allowing them to gain greater understanding of the socio-economic system to be managed.

Deriving Menu Structures through Modal Block Clustering: A Promising Alternative to Hierarchical Techniques

Modal block clustering (MBC) is proposed as an approach more suited to the derivation of menu structures than hierarchical clustering techniques. Problems with the application of hierarchical techniques and pairwise similarity ratings (PWSR) from which the clusters are derived are discussed. MBC defines clusters based on the pattern of common command attributes and provides an objective way to determine the composition and number of menu panels to include in a menu structure. The method also objectively defines command redundancy for the menu panels. The method of MBC was applied to the 97 commands that comprise the CMS operating system resulting in 17 menu categories. The menu categories were used to design a help menu system. The MBC procedure provides a viable methodology for complex systems, such as CMS, which derive increased functionality from numerous command options. System designers can fruitfully and efficiently apply this methodology both to current systems and to proposed systems for which there are no expert users.