{"id":"W4401305302","doi":"10.1016/j.knosys.2024.112312","title":"A safe reinforcement learning algorithm for supervisory control of power plants","year":2024,"lang":"en","type":"article","venue":"Knowledge-Based Systems","topic":"Advanced Control Systems Optimization","field":"Engineering","cited_by":11,"is_retracted":false,"has_abstract":false,"ca_institutions":"Microsoft (Canada)","funders":"Argonne National Laboratory; Laboratory Directed Research and Development","keywords":"Reinforcement learning; Supervisory control; Computer science; Lagrange multiplier; Representation (politics); Control (management); State (computer science); Relaxation (psychology); Control theory (sociology); Optimal control; Mathematical optimization; Power (physics); Lagrangian relaxation; Control engineering; Artificial intelligence; Algorithm; Engineering; Mathematics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004473279,0.0002654801,0.0004929477,0.0002684263,0.00006428522,0.00006378524,0.0001470173,0.0001471616,0.00002114483],"category_scores_gemma":[0.00004392287,0.0002591076,0.0001613995,0.0001975616,0.00002336302,0.0001523679,0.000009503392,0.0001621927,0.00007515163],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002296148,"about_ca_system_score_gemma":0.00008172975,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001314192,"about_ca_topic_score_gemma":0.000003995884,"domain_scores_codex":[0.9984263,0.00008135953,0.0006751094,0.0002633268,0.0001939977,0.0003598692],"domain_scores_gemma":[0.9990194,0.0003998111,0.00007067522,0.0002484249,0.0001702198,0.00009147876],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002559947,0.00001596295,0.00002550144,0.001451536,0.00018713,0.000004380432,0.0003720897,0.9831186,0.004909743,0.0004167104,0.0009322406,0.00854051],"study_design_scores_gemma":[0.001607276,0.0001324263,0.000004977787,0.000831558,0.00003979775,0.000004918255,0.0001603108,0.9436796,0.001125694,0.000006534336,0.05216148,0.0002454351],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0003380863,0.01642686,0.9744368,0.000008022246,0.002525843,0.001288424,0.00008565927,0.0006112477,0.004279059],"genre_scores_gemma":[0.9966328,0.00001748652,0.001105389,0.000005088631,0.000278834,0.000542058,0.00007547121,0.0001044183,0.001238498],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9962947,"threshold_uncertainty_score":0.9999861,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01158204449223214,"score_gpt":0.2260626628308361,"score_spread":0.214480618338604,"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."}}