Multi-source power conditioning: the arbitration problem nobody wants to own

A modern commercial vehicle already has several sources of electrical energy on board. The traction battery. Kinetic energy recoverable at every deceleration. An alternator or range extender. Increasingly, a roof large enough to carry photovoltaic panels. On some duty cycles, shore power between shifts.

In most vehicles these are handled by subsystems that barely speak to each other. Each has its own converter, its own control loop, its own idea of what the battery wants. And when the vehicle turns out to have insufficient range, the design response is almost always the same: fit a bigger battery.

The OMNIA Drive patents take the other route — condition and arbitrate between the sources already present. This article is about why that is a harder problem than it sounds, and about the one number in this field that is routinely oversold.

Why multi-source is genuinely difficult

The naive picture is a DC bus with several things connected to it. The difficulty is that the sources have almost nothing in common.

They differ in impedance and stiffness. A battery is a stiff voltage source with low internal resistance. A photovoltaic array is a current source whose optimal operating point moves continuously with irradiance and cell temperature. A regenerative event is a high-power transient lasting seconds. Connect them naively and they fight: the stiff source absorbs whatever the soft source produces, at whatever the bus voltage happens to be, which is rarely where either wants to operate.

They differ in availability and predictability. Solar output is deterministic over a day and stochastic over a minute. Regenerative energy arrives only when the driver decelerates. Battery energy is available until it is not. A controller that assumes any of them is present will fail on the duty cycle where it is absent.

They differ in control bandwidth by orders of magnitude. Maximum power point tracking for a PV array operates on a timescale of hundreds of milliseconds to seconds. A regenerative braking event needs a response inside tens of milliseconds. Battery state-of-charge estimation is meaningful over minutes. One control architecture has to serve all three without the fast loops destabilising the slow ones.

And they impose conflicting demands on the store. Solar wants to charge whenever the sun is up. Regenerative braking wants to dump a large current with no notice. The battery's health wants neither — it wants moderate current and a moderate average state of charge. If nothing arbitrates, the battery absorbs every conflict, and pays for all of them.

That arbitration is the actual invention. Not the individual converters — those are well-understood engineering — but the logic that decides, continuously, what goes where.

The architecture

The claimed system is a power processing unit conditioning multiple sources, routing traction power against a state-of-charge threshold, and — this is the part that matters — charging the store and driving the motor simultaneously rather than alternating between the two modes.

Five stages: sense, compare, boost, synthesise, deliver.

Sense each source's available power and the store's state. Compare against the threshold and the instantaneous traction demand. Boost or buck each input to the bus operating point. Synthesise a single conditioned supply from whatever is genuinely available in that instant. Deliver to traction and to the store according to the arbitration rule.

The state-of-charge threshold is doing more work than it appears. It is what prevents the two failure modes at either end: routing everything to traction until the store is depleted, or routing everything to the store while traction starves. Making the routing rule a function of SoC rather than of instantaneous power alone is what makes the behaviour stable across a whole duty cycle rather than optimal for one instant.

The number everybody oversells

Vehicle-integrated solar. Here is the honest arithmetic, because the dishonest version has done real damage to this field's credibility.

A 12-metre bus or a large box-body truck has usable roof area on the order of 15–25 m². Commercial modules deliver roughly 180–220 W per m² at standard test conditions. So installed peak capacity is approximately:

20 m² × 200 W/m² ≈ 4 kW_p

In good conditions at Indian latitudes, that array might yield 15–20 kWh per day.

Now the demand side. A heavy commercial vehicle consumes on the order of 100–150 kWh per 100 km. So a full day of excellent sunshine buys somewhere in the region of 10–15 km of range.

That is not a propulsion solution and I will not present it as one. But it is a significant contribution to the auxiliary load — HVAC, refrigeration on a reefer body, telematics, lighting — which on some duty cycles is a substantial fraction of consumption, and which is otherwise drawn from the traction battery at the expense of range. A refrigerated body idling in a depot is exactly the case where 4 kW_p is genuinely useful.

The claim worth making is therefore narrow and specific: vehicle solar meaningfully offsets auxiliary load on duty cycles with high stationary or low-speed time, and contributes marginally to propulsion. Anyone quoting you a large propulsion figure has either not done this arithmetic or is hoping you will not.

Regenerative recovery is the more substantial term. Published measurements on a supercapacitor vehicle report a maximum regenerative energy conversion efficiency of 88% (Zou et al., 2015). That is other people's measurement on other people's hardware — evidence that the effect is real and roughly how large it can be, not a measurement of mine.

The smallest source, described precisely

The specification of this family describes eight generator sub-systems, and among them is ambient RF energy harvesting. It deserves a precise description.

The honest description: Ambient RF harvesting delivers power in the microwatt-to-milliwatt range. It is the lowest-maturity of the eight sources, and its realistic application is powering a sensor node, not contributing to propulsion. It is described because the architecture accepts arbitrary source types, and that generality is worth having.

What this is worth, and to whom

The value of a multi-source architecture is not any single source. It is that the arbitration logic lets you extract more from what is already installed, which is an alternative to adding cells — and cells are the dominant cost, mass and supply-chain constraint in a commercial EV.

The applications where the arithmetic works: high-transient urban duty cycles, vehicles with substantial auxiliary loads, refrigerated and specialist bodies, platforms where mass or battery cost is the binding constraint. The applications where it does not: steady long-haul with a small auxiliary load and a roof already committed to something else.

Family, granted and in force: IN 477613 (6 December 2023) · US 11,117,476 B2 (14 September 2021) · JP 7,109,623 (21 July 2022) · EA 036594 (27 November 2020, validated in Russia and Kazakhstan). Priority 25 September 2015. Each carries a grant number and a grant date in the published register, so the territorial coverage can be checked against the national registers rather than taken on trust.

Measure everything. Validate everything. Improve continuously.

Hemant K. Rohera is an inventor and engineer in Pune, India, sole named inventor on three granted patent families — bioelectronic medical devices, hybrid energy storage and vehicle power electronics. The OMNIA Drive family is available for licensing, assignment or joint development; claim sets, prosecution history and the underlying analysis are available under NDA. ORCID: 0009-0005-3275-1743

References. Zou, Z., Cao, J., Cao, B. & Chen, W. (2015), ISA Transactions, 55, 234–240. The solar arithmetic uses commercial module efficiencies and typical commercial-vehicle consumption figures, stated as ranges.

The complete patent register, with a grant number and grant date for every entry, and the research record, with abstracts and DOIs for the deposited notes, are published at hemant-rohera.vercel.app/patent-register.html and hemant-rohera.vercel.app/research.html.

Search This Blog

Hemant Rohera — Bioelectronics, Energy Storage and Power Electronics