{"id":"W4414603383","doi":"10.1109/tte.2025.3615387","title":"Safe Deep Reinforcement Learning for Energy Management of Electrified Vehicles: Optimal Action Filtering and Battery-in-the-Loop Validation","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Transportation Electrification","topic":"Advanced Battery Technologies Research","field":"Engineering","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"McMaster University","funders":"","keywords":"Reinforcement learning; Powertrain; Benchmark (surveying); Energy management; Metric (unit); Optimal control; Dynamic programming; Minification; Action (physics)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002134175,0.0001555325,0.0001475503,0.0004925446,0.0001449297,0.00002435869,0.0001440038,0.0001167444,0.000007575814],"category_scores_gemma":[0.000003045833,0.0001646356,0.00005441296,0.0007271385,0.00003506961,0.0002220856,5.753142e-8,0.0002606344,8.987976e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001494628,"about_ca_system_score_gemma":0.00001289647,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009894836,"about_ca_topic_score_gemma":0.00002775005,"domain_scores_codex":[0.9988515,0.00003680616,0.0004307973,0.0002419116,0.0001956692,0.0002433626],"domain_scores_gemma":[0.9995058,0.0001438924,0.0000691512,0.0001966868,0.00006681874,0.00001765551],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001177074,0.00003786976,0.00001013137,0.0002096767,0.00005005244,4.699488e-7,0.0001442742,0.6711637,0.2190029,0.0008276675,0.000007366593,0.1084282],"study_design_scores_gemma":[0.0006636994,0.0001278204,0.001101079,0.00006628836,0.00005286408,6.676632e-7,0.0003482883,0.1250072,0.8721009,0.000191063,0.0002109691,0.000129166],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09718182,0.00006703781,0.9017611,0.000128845,0.00007141352,0.0005423254,0.000002973677,0.0001420197,0.0001024949],"genre_scores_gemma":[0.9958264,0.00130951,0.001842772,0.00002207696,0.000005549168,0.0006946891,0.00008239513,0.0000202859,0.0001963394],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8999183,"threshold_uncertainty_score":0.6713644,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01735870452253268,"score_gpt":0.2710587030927269,"score_spread":0.2536999985701942,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}