Dynamic Production Systems Research Articles

Neural and fuzzy model performance evaluation of a dynamic production system

Although mathematical modelling techniques are very well developed, some production processes are difficult to be modelled by these modelling techniques or their math-models are too complex to be used for real-time control due to uncertain, imprecise and vague parameters’ relations. Spray dryers are complex, dynamic and ill-defined production processes. Their product (powder) must have a controllable size distribution consisting of spherical shapes and free-flowing characteristic of particles, which is required for an ideal pressing operation to overcome the product sticking in the dies. The relations of production process' parameters are highly non-linear. In this study, these non-linear parameters were studied and three different soft-computing intelligent models were developed and used to predict uncertain parameter relations. The first is the fuzzy model of the production process; the others are the artificial neural network (ANN) architectures; the back-propagation multilayer perceptron (BPMLP) algorithm and the radial basis function network (RBF). To deal with uncertainty and vagueness of the production system, a method (methodology) based on a fuzzy hierarchical analytic process modelling approach and two ANN approaches was applied. The performance of the BPMLP algorithm was found most vigorous than the RBF and fuzzy modelling approach.

Trans‐Disciplinary Soil Physics Research Critical to Synthesis and Modeling of Agricultural Systems

Synthesis and quantification of disciplinary knowledge at the whole system level, via the process models of agricultural systems, are critical to achieving improved and dynamic management and production systems that address the environmental concerns and global issues of the 21st century. Soil physicists have made significant contributions in this area in the past, and are uniquely capable of making the much‐needed and exciting new contributions. Most of the exciting new research opportunities are trans‐disciplinary, that is, lie on the interfacial boundaries of soil physics and other disciplines, especially in quantifying interactions among soil physical processes, plant and atmospheric processes, and agricultural management practices. Some important knowledge‐gap and cutting‐edge areas of such research are: (1) quantification and modeling the effects of various management practices (e.g., tillage, no‐tillage, crop residues, and rooting patterns) on soil properties and soil–plant–atmosphere processes; (2) the dynamics of soil structure, especially soil cracks and biochannels, and their effects on surface runoff of water and mass, and preferential water and chemical transport to subsurface waters; (3) biophysics of changes in properties and processes at the soil–plant and plant–atmosphere interfaces; (4) modeling contributions of agricultural soils to climate change and effects of climate change on soil environment and agriculture; and (5) physical (cause‐effect) quantification of spatial variability of soil properties and their outcomes, new methods of parameterizing a variable field for field‐scale modeling, and new innovative methods of aggregating output results from plots to fields to larger scales. The current status of the various aspects of these research areas is reviewed briefly. The future challenges are identified that will require both experimental research and development of new concepts, theories, and models.

Youth in the face of vocational agricultural training for sustainable development: whether to push forward or pull back

Vocational training as an inter-temporal investment will bring returns in the future as people responsible take the necessary initiatives to invest today. However, training must be dynamic in nature, especially for agro-entrepreneurship to create structural changes in production, distribution, marketing and consumption. In the preparation of youth to meet future requirements of economic growth, biotechnology, nanotechnology and intelligence in computer technology play prominent roles in the context of knowledge and skill development of the population. In this process, youth should be well informed of conditions and situations of demographics, socio-economics, scientific, energetic and pragmatic aspects of potentials in development through effective vocational training processes in the sphere of sustainable development. By so doing, it would be easy to visualize new production vistas with a new generation of agro-entrepreneurs to yield dynamic production systems for a new world.

Design of practical optimum JIT systems by integration of computer simulation and analysis of variance

This study introduces a framework for re-design of manufacturing systems into practical optimum just-in-time systems by integration of computer simulation and analysis of variance. The conventional JIT approach is mostly applicable to static production systems and the dynamic production systems require a more practical integrated JIT approach. In addition, the re-design of existing dynamic systems into just-in-time systems must follow a practical path, which can be a cumbersome task. This means, a unique practical optimum just-in-time system that considers system's limitations and its dynamic behavior must be designed. To achieve the objective of this study, first, the actual system must be totally modeled and simulated. Second, the integrated simulation model is tested and validated by analysis of variance. Third, the optimum (most fitted) JIT design is developed and tested by modeling actual system's limitations and its dynamic behavior. The framework is applied and tested for an auto production line and a heavy rolling mill workshop.

Optimal Planning of a Multi-Station System with Sojourn Time Constraints

This paper studies a dynamic production system where multiple products must visit stations where inventories are constrained by maximum and minimum sojourn times with neither negative flow nor backlog being allowed. A resource availability constraint limits the aggregate throughput of the stations. The objective is to minimize the sum of flow and inventory cost. The problem is broken down into several single-product serial systems that serve as subroutines of a Lagrangian relaxation routine. This model is implemented in a spreadsheet so that it can be used by the officials of a Chilean institution for planning the operations and defining the optimal allocation of resources.

Receding horizon control of a hybrid production system with deteriorating items

In this paper, a receding horizon control strategy is applied to a dynamic hybrid production system with deteriorating items. Given the current inventory level, we determine the optimal production rates to be implemented at the beginning of each of the following periods over the control horizon. The effectiveness of this approach was in the use of future information of the inventory target level and the desired production rate, which are available. Both the continuous and periodic review policies are investigated. The performances of the proposed control algorithms are illustrated by simulations.

Partnership in a Dynamic Production System

This paper considers two firms that engage in joint production. The prospect of repeated interaction introduces dynamics in that actions that firms take today influence the costliness and effectiveness of actions in the future. Repeated interaction also facilitates the use of informal agreements (relational contracts) that are sustained not by the court system, but by the ongoing value of the relationship. We characterize the optimal relational contract in this dynamic system with double moral hazard. We show that an optimal relational contract has a simple form that does not depend on the past history. The optimal relational contract may require that the firms terminate their relationship with positive probability following poor performance. This may occur even when the firms observe an independent signal for the action of each firm that allows them to assign blame. If, however, the buyer's action does not influence the dynamics, the need for termination is eliminated. The paper applies the method to the issue of sequential versus parallel collaborative product development.

Distribution function of the shortest path in networks of queues

In this paper, we consider acyclic networks of queues as a model to support the design of a dynamic production system. Each service station in the network represents a manufacturing or assembly operation. Only one type of product is produced by the system, but there exist several distinct production processes for manufacturing this product, each one corresponding with a directed path in the network of queues. In each network node, the number of servers in the corresponding service station is either one or infinity. The service time in each station is either exponentially distributed or belongs to a special class of Coxian distribution. Only in the source node, the service system may be modeled by an $M/G/\infty $ queue. The transport times between every pair of service stations are independent random variables with exponential distributions. In method proposed in this paper, the network of queues is transformed into an equivalent stochastic network. Next, we develop a method for approximating the distribution function of the length of the shortest path of the transformed stochastic network, from the source to the sink node. Hence, the method leads to determining the distribution function of the time required to complete a product in this system (called the manufacturing lead time). This is done through solving a system of linear differential equations with non-constant coefficients, which is obtained from a related continuous-time Markov process. The results are verified by simulation.

Unidirectional AGV guidepath network design: A heuristic algorithm

In an AGV system, the design of the guidepath network is one of the most important factors that determine the system effectiveness. Many alternative guidepath design schema have been introduced in the literature. However, the methods generally address the static production environment where the product mix or machine routings are assumed to be stable over time. In today's dynamic production systems with small lot sizes and short product cycles, the assumption of an unchanging product mix over an extended period is unrealistic in some situations. Considerable system inefficiency can be introduced when a network designed under the assumption of a stable product mix is used when indeed the product mix has changed. To avoid this kind of hidden source of inefficiency, what is needed is a tool that can recognize when a previously designed network is inappropriate for a new product mix and then uses the new product routing information to reconfigure a new network appropriate for the new production condition. A heuristic algorithm for the design of AGV guidepath in an environment with changing product mix is proposed here. The algorithm not only makes it possible to adapt the system network as the product mix changes, but also produces new designs at reasonably short computational time.

The Dynamic Production Axioms and System Dynamics Behaviors: The Foundation for Future Integration

Realistically, organizational and/or system performance is dynamic and non-linear. However, in the efficiency literature, system performance is frequently evaluated considering linear combinations of the input/output variables and without explicitly taking into account the causes of efficiency behavior nor the dynamic behavior of systems. Policy decisions based on these results may be sub-optimized because the non-linear relationships among variables, causal relationships, and feedback mechanisms are ignored. This research takes the initial step of evaluating system performance in a dynamic environment, by relating the factors that effect system performance to the policies that govern it. To accomplish this, this paper extends the concepts of the static production axioms into a dynamic realm, where inputs are not instantaneously converted into outputs. The relationships of these new dynamic production axioms to the basic behaviors associated with system dynamics structures are explored.

A simulation model of draught animal power in smallholder farming systems. Part II: Model evaluation and application

Draught Animal Power Simulator (DAP-Simulator) is a computer simulation model developed for smallholder farming systems in which animals are the main source of draught power and nutrition is the main constraint. The description of the model was presented in an earlier serial paper. The objectives of this study were to evaluate the adequacy of the DAP-Simulator model and to demonstrate its application as a decision support system. The predictive performance of the model was evaluated using independent data sets from literature. Linear regression analysis showed ( P<0.05) that force requirements when carting ( r 2=0.72; constant=−0.07 and slope=1.2) and when ploughing ( r 2=0.57; constant=0.01 and slope=0.78), area ploughed per day ( r 2=0.97; constant=0.01 and slope=1.03), daily energy requirements for maintenance ( r 2=0.99; constant=−3.1 and slope=1.09) can be described by this model. The model over-estimated daily weight changes (16%) and also over-estimated feed intake, energy intake and dry matter (DM) digestibility, ranging from 0.4 to 7.8%, for working animals. While the fits obtained were satisfactory, the results emphasised the importance of estimating parameters by more detailed experimentation. Sensitivity analysis was used to examine the response of the model variables to specific changes in the values of the parameters. This is particularly useful as it can be indicative of parameters that management can manipulate so as to increase productivity and reduce costs. The sensitivity analysis indicated that ploughing width (sensitivity index=0.49), ploughing depth (0.58), fraction of soil moved by implement (0.59) and specific soil resistance (0.60) were of limited importance in the estimation of force requirements for ploughing. The force requirements for pulling carts were most sensitive to the weight of the cart (1.76) and the frictional resistance force (1.24). The weight of the load (0.32) and the horizontal component of the pull angle (0.04) had very small effects on the forces for carting loads. It is, therefore, not important to estimate the angle of pull accurately as it has little effect on the output. Area ploughed per day was mainly affected by speed (5), number of passes (4.76), and ploughing width. The most important parameters in simulation of feed intake were fractional rate of passage (3.31), and fractions of degradable (2.5) and indegradable (2.3) fibres. Rate of digestion was most sensitive to fraction of degradable fibre (4.1), with relatively small effects of fractional rates of passage (1.4) and degradation (1.16). A situational analysis of draught power management options and strategies showed that the computer-based simulation model can aid decision-making in complex and dynamic smallholder agricultural production systems.

Performance Modelling of Complex Dynamic Production Systems Using Temporised Object-Oriented Petri Nets with Changeable Structure (TOPNs-CS)

An object-oriented Petri net with changeable structure (OPN-CS) is capable of modelling complex dynamic production systems subject to changes. However, an OPN-CS is limited to functionality modelling of the system. For performance modelling of a complex dynamic production system, OPNs-CS are extended to temporised OPNs-CS (TOPNs-CS). The extension is carried out by temporising OPNs of production resources and the message passing relationships among production resources. The methods of capturing some performance proprieties based on a TOPN-CS model of a system are presented. A case study is provided to show how TCPN-CS is used for modelling complex dynamic production systems. TOPNs-CS can be applied in control, scheduling, and optimisation of the processes of a dynamic production system subject to changes.

Temporized coloured Petri nets with changeable structure (CPN-CS) for performance modelling of dynamic production systems

One of the difficulties in modelling dynamic production systems is the representation of changes to the system along time. Coloured Petri nets with changeable structure (CPN-CS) are capable of modelling systems subject to change. However, this is restricted to functionality modelling of the system. In this paper, we extend CPN-CSs into temporized CPN-CS (TCPN-CS). The extension is carried out by associating relevant modelling elements with time parameters and modifying the mechanism for structure change of the CPN-CS model. TCPN-CS makes it possible to model a dynamic production system considering both functionality and performance. A case study is provided to show that TCPN-CS is effective for modelling dynamic production systems.

EFFICIENT COTTON PRODUCTION IN GREECE WITH THE USE OF GOSSYM - A COTTON CROP SIMULATION MODEL

Cotton (Gossypium hirsutum L.) represents a dynamic crop production system for the agricultural economy of Greece. It is grown in approximately 400,000 ha in average the last few years and supports the income of more than 100,000 rural families. Its production management is based on excessive inputs of water, N fertilizer and agrochemicals, in most cases. This practice is expected to change in the near future due to reformed CAP, the new World trade agreements and serious environmental considerations. New production management strategies should be introduced to optimize yield throughout reduction of major production inputs. A valuable tool to optimise production is represented by GOSSYM, a cotton crop growth and yield simulation model. This model was established, validated and verified in the USA, where it is also used in commercial cotton production. A research study was undertaken in Greece and Spain to evaluate this model under the conditions met in the two cotton producing countries in EU. Results from evaluation in Greece have demonstrated the validity of GOSSYM as a tool for efficient cotton production with low input strategies.

On-line production control using the EVPI approach

This paper deals with the on-line control of a dynamic production system by integrating an off-line nonlinear programming solution with the EVPI (Expected Value of Perfect Information) principle. We consider a production system subjected to random disturbances which has to produce a given target amount by a given due date. There are several possible production speeds to process the target amount, each speed is randomly distributed with a pregiven probability law. The system is observed at discrete points during the course of production. At each such control point, given the observed accumulated amount already produced, the system controller has to set up the speed to be used and to determine the timing of the next control point. The objective is to maximize the expected net profit. The costs considered are: the cost of a single control observation, the penalty cost per unit shortage of the output at the due date and the operating costs per time unit for each speed. The algorithm developed here involves two stages: first we solve the off-line problem which determines the length of time that each speed should be used if there is no control during the course of production. The off-line problem is nonlinear. In order to determine which speed to use first, we apply the EVPI principle. After determining the speed to be used, the system operates with that speed during the corresponding time duration, until the next control point. At this point the actual output is observed and the off-line problem is resolved with the target amount left. The efficiency of the algorithm is evaluated by using extensive simulations.

Valuing the lead time of periodic forecasts in dynamic production systems

The question addressed is, how does the value of climate forecasts change as a function of when they are received? This value is highly sensitive to the expectations of the decision maker at the beginning of the planning horizon. A procedure is developed which allows a decision maker to anticipate the issuance of forecasts in a dynamic programming framework. This procedure is contrasted with a previous methodology in which the decision maker was unaware that a climate forecast would be issued until the forecast was received. Results show that using the ‘unaware’ methodology may overstate the value of receiving forecasts earlier.

System dynamics modelling and simulation of biogas production systems

Design and management of biogas production systems require detailed information on both steady state and dynamic behaviour of the systems. Modelling efforts of steady state and dynamic biogas production systems are reviewed. A dynamic model of biogas production systems based on system dynamics methodology is presented. The model consists of four coupled nonlinear first-order differential equations, two microbial growth equations and the resulting equation for biogas production. The model was programmed in the dynamic simulation language professional DYNAMO suitable for micro-computers. The model's predictive ability is determined by comparing the model output with the observed values reported. Finally, the sensitivity analysis shows that gas production is sensitive to the influent concentration and retention time. It is demonstrated that system dynamics is the simplest and quickest route to the understanding of dynamic systems through models.

Valuing Lead Time of Forecasts in Dynamic Production Systems: Impact of Prior Beliefs

A dynamic production process is analyzed where stochastic events which alter output levels occur during the planning horizon. The impact of the stochastic events can be altered by decisions taken prior to the occurrence of the stochastic events. The question addressed is how does the value of forecasts for the stochastic events change as a function of when they are received? This value is highly sensitive to the expectations of the decision maker. Two sets of expectations are examined. In the first set the decision maker assumes that no forecasts will be forthcoming in the future. In the second set of expectations the decision maker begins the planning horizon knowing that a forecast will become available at a specific time in the future. The results show that the set of expectations on behalf of the producer as to whether a forecast will be forthcoming are very important. In addition, the empirical results show that there is a tradeoff between lead time and accuracy of the forecast such that is can be better to receive a less accurate forecast earlier than a more accurate forecast later.

Valuing Forecast Characteristics in a Dynamic Agricultural Production System

AbstractA dynamic programming model of east‐central Illinois corn production is utilized to determine the value of seasonal climate forecasts. Results indicate that the value of climate forecasts is sensitive to economic conditions as well as forecast characteristics. A trade‐off between forecast accuracy and lead time exists. A less accurate forecast received earlier in the production process may be more valuable than a more accurate forecast received later. Additional characteristics evaluated include forecast periods of greatest importance, prior knowledge assumptions, and interactions between forecasting adjacent periods.

Send-and-Split Method for Minimum-Concave-Cost Network Flows

Many problems from inventory, production and capacity planning, and from network design, exhibit scale economies and can be formulated in terms of finding minimum-additive-concave-cost nonnegative network flows. We reduce the problem to an equivalent one in which the arc flow costs are nonnegative and give a dynamic-programming method, called the send-and-split method, to solve it. The main work of the method entails repeatedly solving set-splitting and minimum-cost-chain problems. In uncapacitated networks with n nodes, a arcs, and d + 1 demand nodes, i.e., nodes with nonzero exogenous demand, the algorithm requires up to n2−13d + s2d operations (additions and comparisons) where s = n log2n + 3a is the number of operations required to solve a minimum-cost-chain problem with nonnegative arc costs on the augmented graph formed by appending a node and an arc thereto from each node in the graph. If also the network is k-planar, i.e., the graph is planar with all demand nodes lying on the boundary of k faces, the method requires at most n2−kd3k + sd2k operations. The algorithm can be applied to capacitated networks because they can be reduced to equivalent uncapacitated ones. These results unify, significantly generalize (e.g., to cyclic problems), and sometimes improve upon (e.g., for tandem facilities) known polynomial-time dynamic-programming algorithms for Wagner and Whitin's (Wagner, H. M., Whitin, T. M. 1958. Dynamic version of the economic lot size model. Management Sci. 5 89–96.) dynamic economic-order-quantity problem, Zangwill's (Zangwill, W. I. 1969. A backlogging model and a multi-echelon model of a dynamic economic lot size production system—A network approach. Management Sci. 15 506–527.) generalization to tandem facilities, and Veinott's (Veinott, Jr., A. F. 1969. Minimum concave-cost solution of Leontief substitution models of multi-facility inventory systems. Oper. Res. 17 262–291.) increasing-capacity warehousing problem. The networks for the finite-period versions of these problems are each 1-planar. The method improves upon Zangwill's (Zangwill, W. I. 1968. Minimum concave cost flows in certain networks. Management Sci. 14 429–450.) related O(and) running-time dynamic-programming method for finding minimum-additive-concave-cost non-negative flows in circuitless single-source networks. We also implement the method to solve in polynomial time the (d + 1)-demand-node and k-planar versions of the minimum-cost forest and Steiner problems in graphs. The running time for the (d + 1)-demand-node Steiner problem in graphs is comparable to that of Dreyfus and Wagner's (Dreyfus, S. E., Wagner, R. A. 1971. The Steiner problem in graphs. Networks 1 195–207.) method.