What is FCS-MPC for a bidirectional DC-DC converter?

Finite Control Set Model Predictive Control (FCS-MPC) predicts the future behaviour of the converter states over a short horizon and selects the optimal switching state from the discrete finite set that minimises a cost function, delivering faster dynamics and constraint handling compared with classical PI control.

Which software is used for the Bidirectional DC-DC Converter project?

MATLAB R2024b / Simulink with Simscape Electrical, discrete PWM generators, state-space models of Buck and Boost modes, and custom FCS-MPC cost-function scripts.

Bidirectional DC-DC Converter MATLAB Simulink | FCS-MPC Battery Microgrid

Bidirectional DC-DC Converter with FCS-MPC for DC Microgrid Storage

Storage units are critical in microgrids to ensure stable operation when renewable sources are intermittent. This project implements a finite control set model predictive control (FCS-MPC) strategy to manage charging and discharging of a battery through a bidirectional DC-DC converter. Unlike traditional PI controllers, FCS-MPC optimises switching actions in real time, respects system constraints and delivers superior dynamic performance while continuously supplying a 15 W DC load.

Key Technical Highlights

Bidirectional Buck / Boost topology
State-space models for both modes
FCS-MPC cost-function optimisation
SOC-based energy management (10–90 %)
PI controller baseline comparison
MATLAB / Simulink validation
15 V / 15 W continuous load supply
10 kHz switching frequency

MMATLAB R2024b SSimulink ESimscape Electrical PFCS-MPC BBattery SOC

Proposed Power System Topology

DC voltage source + 2.5 Ah battery + 15 W load form a DC microgrid. The bidirectional converter is the key interface that transfers energy between source, battery and load.

DC Source
V_DC = 15 V

→

Bidirectional
DC-DC Converter

→

Battery
6 V / 2.5 Ah

→

15 W Load
R = 15 Ω

Paper title and abstract — FCS-MPC bidirectional converter

Source paper: A New Advanced Strategy for Controlling the Charging and Discharging of a Storage Unit in a Microgrid Using a Finite Control Set Predictive Model

Figure 1 — Proposed power system topology (DC source, bidirectional converter, battery, load)

Bidirectional DC-DC Converter — Buck & Boost Modes

The converter operates in Boost mode when the battery discharges to the load and in Buck mode when the DC source charges the battery while still feeding the load.

⬆

Boost (Discharge)

Battery is energy source; converter steps voltage up to supply the 15 V load. SOC decreases toward 10 % floor.

⬇

Buck (Charge)

DC source charges the battery and supplies the load. Converter steps voltage down; SOC rises toward 90 % ceiling.

⚡

Mode Decision

SOC compared with SOCinit = 50 %. Below → Buck charge; above → Boost discharge. Protects battery lifetime.

📐

State-Space Model

Continuous-time models for both modes feed the FCS-MPC predictor; discrete finite set of switching states is evaluated each sample.

Figure 2 — Bidirectional DC/DC converter circuit (shared inductor, dual switches, filter capacitors)

Finite Control Set Model Predictive Control (FCS-MPC)

At every sampling instant the controller predicts the future inductor current and output voltage for each admissible switching state, evaluates a cost function that penalises tracking error and switching effort, and applies the optimal state for the next period.

Why FCS-MPC outperforms PI

Classical PI controllers require separate tuning for Buck and Boost, suffer from limited bandwidth under large-signal disturbances and cannot explicitly enforce constraints. FCS-MPC evaluates the discrete finite set of switch combinations, includes current and voltage limits inside the cost function, and yields faster settling with lower overshoot when the renewable source or load changes abruptly.

Real-time optimisation of switching states
Explicit constraint handling (current, voltage, SOC)
Single controller structure for both modes
Superior transient response vs PI
Higher conversion efficiency under test scenarios

MATLAB / Simulink Implementation

Complete energy chain and control chain are modelled in Simulink. A mode switch selects the correct state-space matrices; the FCS-MPC block computes the optimal duty/switch command; a parallel PI path provides the baseline comparison.

Matlab/Simulink simulation of bidirectional DC-DC converter

Figure 8 — MATLAB/Simulink simulation of the bidirectional DC/DC converter energy chain

Figure 9 — FCS-MPC controller simulation topology

Figures 10–11 — PI controller path and SOC-based energy management logic

Study System Parameters

Table 1 — Study System Parameters (from paper)

Parameter	Value	Unit
V_DC	15	V
i_DC	1	A
V_batt	6	V
C_batt	2.5	Ah
R (load)	15	Ω
L	1	mH
C_b	20	µF
C_l	500	µF
r_L	0.2	Ω
r_C	0.001	Ω
f_s	10	kHz

Simulation Results — FCS-MPC vs PI

Buck (charge) and Boost (discharge) scenarios demonstrate continuous 15 W load supply, battery current/voltage trajectories and the superior tracking of FCS-MPC.

Buck mode battery voltage and current FCS-MPC vs PI

Buck mode (battery charging) — battery current and voltage under FCS-MPC and PI

Boost mode (battery discharging) and comparative performance metrics

Continuous Load Supply

15 W load remains powered in both charge and discharge regimes; converter seamlessly transitions between Buck and Boost.

Faster Dynamics

FCS-MPC settles inductor current and output voltage faster than PI under step changes of SOC or source availability.

Constraint Awareness

Current and voltage limits are embedded in the cost function, preventing over-current during mode transitions.

Higher Efficiency

Paper results show FCS-MPC delivers higher conversion efficiency and better overall performance than the PI baseline.

Sample MATLAB Code & Parameter Scripts

Representative excerpts for system parameters, state-space matrices and a simplified FCS-MPC cost evaluation loop.

bidirectional_dc_dc_params.m — System Parameters & Mode Flags

% bidirectional_dc_dc_params.m — Parameters for FCS-MPC Bidirectional Converter clc; clear; close all; % DC source & battery Vdc = 15; % V Vbatt = 6; % V nominal Cbatt = 2.5; % Ah SOC_init = 0.50; SOC_min = 0.10; SOC_max = 0.90; % Power stage L = 1e-3; % H Cb = 20e-6; % F Cl = 500e-6; % F rL = 0.2; % Ohm rC = 0.001; Rload = 15; % Ohm → 15 W at 15 V fs = 10e3; % Hz Ts = 1/fs; % Mode flag: 1 = Boost (discharge), 0 = Buck (charge) mode = 1; % set by SOC management fprintf('[✓] Bidirectional DC-DC params loaded | fs = %.0f kHz | Rload = %.0f Ω\n', fs/1e3, Rload);

fcs_mpc_step.m — One FCS-MPC Prediction & Cost Evaluation (conceptual)

% fcs_mpc_step.m — Finite Control Set MPC for Bidirectional Converter % Evaluates cost for each admissible switching state S = {0,1} function S_opt = fcs_mpc_step(iL, vO, iL_ref, vO_ref, mode, params) S_set = [0, 1]; % finite control set J_min = inf; S_opt = 0; for S = S_set % Discrete prediction using state-space of selected mode [iL_p, vO_p] = predict_state(iL, vO, S, mode, params); % Cost: tracking error + optional switching penalty J = params.lambda_i * (iL_p - iL_ref)^2 + ... params.lambda_v * (vO_p - vO_ref)^2; if J < J_min J_min = J; S_opt = S; end end end % In Simulink: call every Ts; apply S_opt to the bridge switches

Command Window — Typical Simulation Summary

>> run_bidi_fcs_mpc_sim
[*] Compiling Simscape Electrical bidirectional topology... [✓] Buck & Boost state-space matrices loaded [✓] FCS-MPC cost weights: λi = 1.0, λv = 0.5 ──────────────────────────────────────────────────────── [BUCK MODE] Battery charging, SOC 10 % → 90 % Load power continuous: 15.0 W [BOOST MODE] Battery discharging, SOC 90 % → 10 % Load power continuous: 15.0 W ──────────────────────────────────────────────────────── [✓] FCS-MPC settling faster than PI; efficiency higher under test scenarios [✓] IEEE-format waveforms and tables exported for report

Advantages of the Bidirectional DC-DC FCS-MPC Project

Modern Predictive Control

Replaces classical PI with finite-set model predictive control — a hot research topic in power electronics and microgrids.

Full Buck + Boost Coverage

Single topology and controller handle both charging and discharging with SOC-based mode selection.

Constraint-Aware Design

Current, voltage and SOC limits are explicitly included in the cost function, improving safety and battery life.

Clear PI Baseline

Side-by-side comparison with a well-tuned PI controller makes the performance gain measurable and viva-ready.

Simulink + Script Ready

Deliverable includes .SLX models, parameter .m files, cost-function scripts and waveform capture scripts.

IEEE / University Report

Complete documentation, calculation tables, result figures and viva Q&A tailored to VTU / Anna / NIT formats.

Frequently Asked Questions

Finite Control Set Model Predictive Control predicts the future inductor current and capacitor voltage for each admissible switching state, evaluates a cost function, and applies the optimal state. It handles constraints naturally and reacts faster than classical PI under large disturbances.

MATLAB R2024b, Simulink, Simscape Electrical, discrete-time state-space models for Buck and Boost modes, and custom FCS-MPC cost-function scripts. A parallel PI path is retained for comparison.

Yes. Full .SLX models, MATLAB .m parameter and controller scripts, parameter tables, waveform results, university-format IEEE report, presentation PPT and viva defence Q&A.

A 15 W / 15 V resistive load and a 6 V, 2.5 Ah battery. Switching frequency is 10 kHz; inductor 1 mH; filter capacitors 20 µF and 500 µF. SOC is managed between 10 % and 90 %.

Project Deliverables & Support

Everything needed from project inception to final viva defence.

Simulink Models
.SLX Files

Energy + Control

MATLAB Scripts
.M Files

Params + FCS-MPC

Circuit
Diagrams

Buck / Boost

Project
Report

IEEE Format

PPT
Slides

25+ Slides

Waveform
Results

FCS-MPC vs PI

Parameter
Tables

L, C, SOC

Viva
Q&A

60+ Questions

Expert
Support

WhatsApp / Zoom

PhD
Extension

Research Ready