This landscape identifies 18 published patent families filed between 2015 and mid-2026, using a search combining bidirectional converter hardware terms with simulation, average-model, electro-thermal, and digital-twin language restricted to relevant IPC classes. That is a small, specialized set compared to the broader bidirectional converter hardware category, reflecting how narrow the modeling-specific overlap actually is. The true 2025-2026 count is likely higher once publication catches up, since patent publication typically lags filing by around 18 months.

The assignee base is a long tail rather than a concentrated leaderboard: universities such as Harbin Engineering University and Southwest Jiaotong University appear alongside Chinese state grid research entities, with isolated single filings from firms like BMW and SK hynix. No single organization holds a commanding share of the 18 tracked families. Recent-year momentum data shows most named assignees at zero filings in the latest year, which is more likely a publication-lag artifact than a sign the field has gone quiet.

An average model is a simplified mathematical representation that captures a converter’s steady-state and slow dynamic behavior for control design and simulation, without modeling every switching event. A digital twin, as described in filings like WO2026157018A1, goes further: it pairs a physical converter with real-time data sampling, a control system, and a synchronized virtual model — often FPGA-accelerated — for continuous monitoring, fault diagnosis, and lifecycle health management. In this patent set, average-model and electro-thermal modeling claims are older and more established, while digital-twin system claims are the newer, more actively filed category.