Non-linear economic model predictive control of water distribution networks

Abstract

This paper addresses a non-linear economic model predictive control (EMPC) strategy for water distribution networks (WDNs). A WDN could be considered as a non-linear system described by differential-algebraic equations (DAEs) when flow and hydraulic head equations are considered. As in other process industries, the main operational goal of WDNs is the minimisation of the economic costs associated to pumping and water treatment, while guaranteeing water supply with required flows and pressures at all the control/demand nodes in the network. Other operational goals related to safety and reliability are usually sought. From a control point of view, EMPC is a suitable control strategy for WDNs since the optimal operation of the network can not be established a priori by fixing reference volumes in the tanks. Alternatively, the EMPC strategy should determine the optimal filling/emptying sequence of the tanks taking into account that electricity price varies between day and night and that the demand also follows a 24-hour repetitive pattern. On the other hand, as a result of the ON/OFF operation of parallel pumps in pumping stations, a two-layer control scheme has been used: a non-linear EMPC strategy with hourly control interval is chosen in the upper layer and a pump scheduling approach with one-minute sampling time in the lower layer. Finally, closed-loop simulation results of applying the proposed control strategy to the D-Town water network are shown.

Receding Horizon Control for Drinking Water Networks: The Case for Geometric Programming

“…The study [2] obtains optimal management strategies in urban water cycle via MPC. The authors in [3] address a nonlinear economic MPC strategy to minimize the economic costs associated with pumping and water treatment. A nonlinear MPC controller is designed in [4] to meet consumer demands at desired pressures.…”

“…2) Junctions and Pipes: Junctions are the points where water flow merges or splits. The expression of mass conservation of the i th junction at time k can be written as (3) in Tab. II, and d i stands for end-user demand that is extracted from node i.…”

Geometric Programming-Based Control for Nonlinear, DAE-Constrained Water Distribution Networks

“…The pump properties decide the relationship function between pump flow and head increase [17]. Generally, it can be expressed as (7), where h M ij is head increase of the pump from i to j; h 0 is the shutoff head for the pump; q ij is the flow through a pump; s ij ∈ [0, 1] is the relative speed of the same pump; r and ν are the pump curve coefficients. We also assume that reservoirs have infinite water supply, the head of the i-th reservoir h R i is fixed, and this can be viewed as an operational constraint.…”

Minimizing the Effect of Substantial Perturbations in Military Water Systems for Increased Resilience and Efficiency

“…An excellent feature article was published in 2002 providing an overview of feedback control as it applies to water quality [15]. A couple of key papers expand on the application of MPC to water systems by making improvements to the controllers or reducing uncertainty in the algorithms [16][17][18].…”