Przemyslaw Ignaciuk

Discrete-Time Control of Production-Inventory Systems With Deteriorating Stock and Unreliable Supplies

This paper addresses the problem of establishing an efficient (fast yet robust) inventory management policy for production-inventory systems in which the stock deteriorates with time and the replenishment process is unreliable. The stock used to satisfy uncertain, variable demand is refilled using multiple supply options. Owing to information distortion, product defects, manufacturing faults, or improper transportation, the goods quantity obtained from the suppliers differs from the ordered one. The setting is modeled as an uncertain, time-varying discrete-time system with multiple time delays. A new compensation mechanism, which provides a smooth ordering pattern and ensures closed-loop stability for a delay spanning arbitrary number of review periods is proposed. The developed strategy outperforms the classical order-up-to policy in mitigating demand variations and prevents the bullwhip effect despite unpredictable distortions of the supply process.

LQ Optimal and Robust Control of Perishable Inventory Systems With Multiple Supply Options

In this technical note, an analytical framework for studying the dynamics of perishable inventory systems is developed using control-theoretic approach. In the considered systems, the deteriorating stock at a goods distribution center is used to fulfill unknown, time-varying demand. The stock is replenished using multiple supply options characterized by different lead-time delay. In contrast to the classical stochastic, or heuristic approaches, a formal design procedure based on discrete-time linear-quadratic (LQ) optimal control is employed. The analytically derived suboptimal controller satisfies positivity constraints and ensures full demand satisfaction for any bounded demand pattern. Moreover, a methodology for analysis of parametric uncertainties is elaborated, and another, nonlinear controller that guarantees robustness to uncertain delay variations is designed.

Nonlinear Inventory Control With Discrete Sliding Modes in Systems With Uncertain Delay

In this paper, discrete sliding-mode (DSM) control concepts are applied to develop an efficient control algorithm for a class of uncertain systems with delay-logistic networks. In the analyzed systems, the stock used to satisfy unknown, variable demand placed at a goods distribution center is replenished with delay from a number of distant supply sources. In the considered setting, the order quantity is fixed and the instants of placing the order serve as the decision variable, which perfectly suits the switching input signals in DSM control systems. It is proved that, with the proposed nonlinear algorithm applied, the stock level does not exceed the assigned warehouse space. Moreover, it is also shown that the stock is never depleted, which ensures full demand satisfaction and maximum service level. Comparison with the classical inventory policy is provided, and improved performance of the DSM algorithm is demonstrated.

Linear-quadratic optimization in discrete-time control of dynamic plants operated with delay over communication network

The paper addresses the problem of remote plant control in networked environment with the use of linear-quadratic (LQ) optimal controller. Since information exchange in networked control systems is essentially a discrete-time process, the controller design is conducted directly in discrete time domain. It is formally shown that the optimization problem for systems involving delay may be reduced to an equivalent delay-free problem with variables expressed in delay compensator dynamics. As a result, the computational complexity to solve the matrix design equations is reduced. The robustness of optimal control system is evaluated analytically and its properties illustrated experimentally.

Network nodes play a game – a routing alternative in multihop ad-hoc environments ☆

Abstract A key aspect to consider in delivering data from one node to another in a communication network is finding a suitable path through the relaying nodes, i.e., routing. The standard routing solutions work well in the reliable networks, but do not achieve acceptable performance in low-power, noisy, and dynamic environments due to the high overhead associated with establishing a consistent view of the network. The paper presents a different solution to the problem of data delivery based on an evolutionary game between the network nodes. In the proposed approach, instead of forming a fixed path, the nodes take independent forwarding decisions. The set of nodes participating in the data transfer self-adjusts to the current networking conditions by playing the game, without involvement of a control plane. As a result, a low-overhead alternative to routing is obtained. The proposed solution is particularly well suited for the ad-hoc wireless networks serving time-sensitive data, e.g., it can provide an efficient communication platform for low-cost networked control systems.

Inventory management in systems with perishable goods and unreliable production-delivery channels

The paper considers the problem of establishing an efficient control strategy for production-inventory systems in which the stock replenishment process is unreliable. In the analyzed setting, the stock used to satisfy unknown, variable demand deteriorates with time. It is refilled using multiple production-delivery channels. Due to information distortion, product defects, or improper transportation, the shipments may arrive damaged, or incomplete. The setting is modeled as a time-varying discrete-time system with multiple time-delays. A new delay compensation mechanism, which provides smooth ordering pattern and ensures closed-loop stability for arbitrary delay, is proposed. The control system properties are formally proved and illustrated in numerical experiments.

Control of production-inventory systems with deteriorating stock and unreliable delayed supply channel

The paper considers the problem of establishing an efficient control strategy for production-inventory systems in which the stock replenishment process is unreliable. In the analyzed setting, the stock used to satisfy unknown, variable demand deteriorates with time. It is refilled with delay from a remote supply source. Due to information distortion, production, or transportation faults, the shipments may arrive damaged, or incomplete. The setting is modeled as a time-varying discrete-time system with delay. A new dead-time compensation mechanism, which provides smooth ordering pattern and ensures closed-loop stability for arbitrary delay, is proposed. The properties of the proposed control system are formally proved and illustrated in numerical experiments.

On Implementation of Energy-Aware MPTCP Scheduler

The majority of networking devices currently available are equipped with a few communication interfaces. However, in most cases, only one of them is used for data transfer as dictated by the design premises of the fundamental transport protocol – TCP. Recently, a variant of TCP – Multipath TCP (MPTCP) – that allows for simultaneous transmission over different paths has been developed. While it allows for spreading the data stream among different interfaces and paths, the way it is actually performed has been left for further investigation. The stream splitting can be optimized according to different quality indicators. The paper discusses the implementation properties of an energy-aware load balancing method tested in a physical networking environment. A comparison with the reference solution is also provided.

State-Space Modeling and Analysis of Order-up-to Goods Distribution Networks with Variable Demand and Positive Lead Time

The paper addresses inventory management problem in goods distribution systems organized in a network structure. Contrary to majority of earlier approaches, that restrict the analysis to chain and tree-like configurations, mesh connectivity is considered. The external, a priori unknown demand may be imposed on any node of the managed network and goods trans-shipments are effectuated with positive lead time. The paper proposes a state-space model of the network node interaction under distributed order-up-to inventory policy. Managerial insights regarding target stock and warehouse space selection are provided and supported with numerical tests.

Energy Efficient Dynamic Load Balancing in Multipath TCP for Mobile Devices

Over the last few years, one can observe a shift in preferences to use wireless media, including WiFi and cellular, rather than wired technologies for the communication purposes. On the other hand, a variant of TCP that allows for simultaneous transmission over different paths has been developed. This paper addresses the problem of optimal load distribution among the interfaces available at a networking device, e.g. mobile phone, with the objective of minimizing the overall energy consumption. A dynamic allocation algorithm, adapting to current, time-varying channel capacity is designed. Comparison with earlier approaches is provided and superiority of the proposed solution is demonstrated in numerical tests.