Virtual Power Plant AI and Machine Learning Patent Landscape 2026
- 17 families, one clear leader. Power Management Holdings (U.S.), Inc. holds the most-cited dispatch-optimization filings, including the newly granted US12462313B2.
- Filing peaked in 2023 at 6, not in the most recent year. 2022 shows just 1 filing and the 2026 count is partial, so the apparent slowdown is largely a publication-lag artifact.
- AI-specific claims are still thin. Only 3 of 17 records sit in G06N (AI-model computing) — most of the field claims dispatch as a business-method-plus-grid-control combination instead.
Top-5 share is the combined record count of the five largest assignees divided by all 17 records in scope (CR5), not by the ranked leaders only.
A small, business-method-heavy field still forming its claim structure
Virtual power plant AI and machine learning patenting is a narrow field by patent-landscape standards: 17 patent families cover the intersection of VPP aggregation with load forecasting, dispatch optimization and reinforcement-learning control. The technology composition skews toward G06Q business-method claims paired with H02J grid-control claims rather than toward claims on the machine-learning model itself — G06N AI-model filings account for only 3 of the 17 records. Filing activity peaked in 2023 and has not yet re-accelerated, though the most recent years are understated by publication lag.
Power Management Holdings (U.S.), Inc. is the clearest commercial anchor in the dataset, holding both the most-cited older dispatch patent and the newest grant, US12462313B2. The rest of the corpus is fragmented across Chinese university and grid-utility filers and a scatter of single-family entrants across Canada, India, Australia and Europe — a structure that leaves several technical sub-areas, notably EV-fleet dispatch and reinforcement-learning-based allocation, comparatively open.
Filing trend and technology composition
Seventeen patent families make this a small, still-forming corpus. The trend and IPC split below show where filing activity has actually landed, not where the technology could theoretically go.
Filing activity, 2017–2026
Filings stayed at zero through 2017, then built to a peak of 6 in 2023 before easing back — 2022 sits at just 1, and the 2026 figure is a partial year, so read the recent drop as incomplete data rather than a confirmed decline.
IPC subclass distribution
G06Q (business/commerce data processing) and H02J (power supply and grid systems) dominate at 15 and 13 records respectively, confirming that most claims combine a dispatch/optimization business method with an actual grid-control step. G06N (AI models), G05B (control systems), G06F and B60L each appear in single digits, marking them as the less-claimed edges of the field.
Shares are the percentage of the 17 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Virtual Power Plant AI and Machine Learning with Eureka
This page is one run against one query. Ask Eureka your own question about virtual power plant ai and machine learning and every answer comes back with the patent numbers behind it.
Try EurekaMost-cited and most recent records
Energy dispatch optimization using a fleet of distributed energy resources
A dispatch optimization system and virtual power plant can be utilized and controlled in order to support the operations of a power distribution system. Upon determining an electrical need, the system allocates the energy adjustment among distributed energy resources of a fleet, basing the allocation on the economic costs and storage costs of using each resource.Granted to Power Management Holdings (U.S.), Inc. on 2025-11-04 — the most recent grant in the corpus.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | CN115271467A | 考虑电碳协同优化的虚拟电厂调度优化方法及应用 | 31 |
| 2 | CN120474103A | 基于多智能体的虚拟电厂智能控制方法及系统 | 11 |
| 3 | US10635056B2 | Model and control virtual power plant performance | 4 |
| 4 | US20240127370A1 | Energy dispatch optimization using a fleet of distributed energy resources | 3 |
| 5 | CN121076815A | 基于深度强化学习的虚拟电厂协同优化调度方法及系统 | 2 |
| 6 | CA2846342A1 | Use of demand response (DR) and distributed energy resources (DER) to mitigate the impact of variable energy … | 2 |
Ranked by citation count within this search; older filings naturally accumulate more citations, so treat this as an influence signal rather than a current-relevance ranking.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. Publication numbers are shown where the record carries one (6 of 6 rows); clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the filing pattern signals
With only 17 families in the corpus, every structural signal — citation concentration, IPC split, geographic spread — carries more weight than it would in a larger field. Here is what stands out.
Influence sits on one carbon-dispatch filing
CN115271467A, an electricity-carbon co-optimization dispatch method, draws far more citations than anything else in the set. That is a strong influence signal inside this corpus, but it is also an older-record effect — citation counts favour filings with more time on record, not necessarily the most commercially active current approach.
Business-method claims outnumber pure AI claims
G06Q (business/commerce processing) leads at 15 records, ahead of H02J (grid systems) at 13, with G06N (AI models) trailing at just 3. Most filers are claiming the optimization logic as a business method paired with a grid-control step, rather than claiming a distinct machine-learning model as the invention.
No filer shows fresh-year activity
Every assignee in the recent-momentum list, from Power Management Holdings to multiple Chinese State Grid research institutes, shows 0 filings in the latest tracked year. Given the roughly 18-month publication lag, this reads as an incomplete data window rather than an actual pause in R&D.
Filing is split across offices, not concentrated in one
China leads with 5 records, the US follows with 4, and Canada holds 3 — a spread that suggests no single jurisdiction has become the default filing venue for VPP AI/ML claims yet. Europe and Australia each show just 1, marking them as comparatively open filing venues.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to virtual power plant ai and machine learning, with the prior art for and against each one.
| Assignee | Co-assignee | Shared families |
|---|---|---|
| POWER MANAGEMENT HOLDINGS US INC | RUTH MICHAEL K | 1 |
| POWER MANAGEMENT HOLDINGS US INC | GASSNER ANDREW R | 1 |
| POWER MANAGEMENT HOLDINGS US INC | BROWN DUNCAN R | 1 |
| Zhejiang University | State Grid Qinghai Electric Power Company Economic and Technical Research Institute | 1 |
| Zhejiang University | State Grid Qinghai Electric Power Company Clean Energy Development Research Institute | 1 |
| Zhejiang University | North China Electric Power University | 1 |
| State Grid Qinghai Electric Power Company Economic and Technical Research Institute | State Grid Qinghai Electric Power Company Clean Energy Development Research Institute | 1 |
| State Grid Qinghai Electric Power Company Economic and Technical Research Institute | North China Electric Power University | 1 |
Ten co-assignee pairs exist in the corpus, all at a single-pair strength; the strongest three all link Power Management Holdings (U.S.), Inc. with individual named inventors (Ruth Michael K, Gassner Andrew R, Brown Duncan R), not with other companies. There is no visible multi-company alliance structure in this field yet.
Who holds the claims, and where the gaps are
With only 17 families in play, the field has one entrenched filer and a long tail of single-family entrants — mostly Chinese universities and grid-utility research institutes — none of whom show renewed filing activity in the latest tracked year.
Power Management Holdings (U.S.), Inc.
Holds the corpus's most-cited US dispatch patent and its newest grant, both built around fleet-level cost-based allocation. Its co-assignee pairs link only to named individual inventors, not to other companies, suggesting the claim position is held tightly rather than licensed out.
State Grid research institutes and universities
Zhejiang University and several State Grid provincial research and economic-technology institutes each hold single-family positions, mostly in dispatch and control methods. None shows recent-year momentum, and none has yet built a multi-generation family the way Power Management Holdings has.
Thin but present filing spread
Canada's 3 records and India's 2 show some non-US, non-China filing interest, while Australia and Europe sit at just 1 record each. This spread indicates the field has not consolidated around a single filing jurisdiction, leaving smaller offices comparatively open.
| Assignee | Recent year | YoY |
|---|---|---|
| POWER MANAGEMENT HOLDINGS US INC | 0 | — |
| OPEN ACCESS TECH INT | 0 | — |
| Zhejiang University | 0 | — |
| State Grid Qinghai Electric Power Company Economic and Technical Research Institute | 0 | — |
| State Grid Qinghai Electric Power Company Clean Energy Development Research Institute | 0 | — |
| State Grid Zhejiang Electric Power Co., Ltd. Chun'an County Power Supply Company | 0 | -100% |
| State Grid Zhejiang Electric Power Co., Ltd. Hangzhou Power Supply Company | 0 | -100% |
| State Grid Zhejiang Electric Power Research Institute Co., Ltd. | 0 | -100% |
Where to take this analysis
The 17-family corpus points to specific next actions depending on whether you are filing, licensing or tracking competitors.
Check claim overlap before drafting a dispatch-optimization filing
Any cost-based fleet allocation claim should be checked against Power Management Holdings' two-generation family (US10635056B2 and US12462313B2) before drafting, since that is the densest claim territory in the corpus.
Run a claim comparison in EurekaTrack the under-claimed EV-fleet and RL-allocation routes
EV-fleet-as-VPP-resource claims and reinforcement-learning-based cost allocation both show minimal filing density, making them worth monitoring for early competitor activity before the space fills in.
Set up a monitoring alert in EurekaRevisit the 2025–2026 filing count once publication lag clears
The apparent drop-off after 2023 is likely a data-window artifact given the roughly 18-month publication lag; re-pull the trend in six to twelve months for a truer read on recent momentum.
Explore the full trend data in EurekaQuestions practitioners ask about VPP AI patents
The tracked corpus for this specific search — virtual power plants combined with load forecasting, dispatch optimization or reinforcement-learning control, filtered to the relevant AI and grid IPC codes — contains 17 patent families. That is a small, still-forming field rather than a mature one. Broader VPP filings that do not specify an AI/ML control method will not appear in this count, so treat 17 as the AI-specific slice, not the whole VPP patent landscape.
Power Management Holdings (U.S.), Inc. holds the most-cited US filings in the set, including the newly granted US12462313B2 and the earlier US10635056B2, both centred on fleet-level dispatch allocation. Beyond that filer, the field is fragmented: Chinese university and grid-utility filers such as Zhejiang University and several State Grid research institutes appear, but recent-year momentum for all of them, including Power Management Holdings, shows 0 new filings in the latest tracked year. That flatness reflects publication lag as much as any real slowdown.
US12462313B2, granted to Power Management Holdings (U.S.), Inc. on 2025-11-04, covers a dispatch optimization method that allocates an energy adjustment across a fleet of distributed energy resources based on both economic cost and storage cost, in response to a detected electrical need on a power distribution system. It does not cover forecasting-only systems or single-variable allocation methods that ignore either cost factor. Because it only just granted, its practical claim boundaries are still being tested against later filings.
Yes, but sparingly relative to the rest of the field — G06N (AI-model computing) appears in only 3 of the 17 tracked records, including CN121076815A, which applies deep reinforcement learning to VPP coordinated dispatch, and CN120474103A, which uses a multi-agent control architecture. Most of the corpus instead frames dispatch as an optimization or allocation problem under G06Q and H02J rather than an explicitly learned-policy approach. That gap suggests RL-based control is still an open, lightly-claimed sub-area rather than the field's default method.
The densest overlap risk is in fleet-level dispatch allocation that ranks resources by cost, since Power Management Holdings' family (US10635056B2 and US12462313B2) and the highly-cited CN115271467A already occupy that exact claim space. Lower-risk, less-claimed areas include electric-vehicle-fleet-as-VPP-resource claims (only 1 record under B60L) and reinforcement-learning-policy-based allocation, which uses a different decision mechanism than the cost-based claims already on file. Checking the receiving-office spread also matters: Europe has only 1 filing in this set, so EPO coverage is comparatively open regardless of the underlying method.
Research Virtual Power Plant AI and Machine Learning in depth with Eureka
Go past this page: query the whole virtual power plant ai and machine learning corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.