{"id":"W4391214730","doi":"10.2139/ssrn.4706893","title":"Deep Reinforcement Learning for Economic Battery Dispatch: A Comprehensive Comparison of Algorithms and Experiment Design Choices","year":2024,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Smart Grid Energy Management","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University","funders":"","keywords":"Reinforcement learning; Hyperparameter; Computer science; Battery (electricity); Benchmark (surveying); Economic dispatch; Artificial intelligence; Machine learning; Power (physics); Electric power system","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0266127,0.001108115,0.001889371,0.0005468685,0.0003697096,0.001186393,0.001482721,0.001867744,0.002868059],"category_scores_gemma":[0.05155369,0.0006475846,0.0009850407,0.0004749761,0.001091088,0.002600301,0.00125762,0.002182434,0.0002754022],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001645021,"about_ca_system_score_gemma":0.002222223,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002022093,"about_ca_topic_score_gemma":0.001702119,"domain_scores_codex":[0.9918153,0.006683296,0.0002577017,0.0005261308,0.0004685905,0.0002490117],"domain_scores_gemma":[0.9233214,0.069457,0.001466021,0.003560709,0.001608797,0.0005861489],"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":[0.01041789,0.002116575,0.004937813,0.0005537259,0.0008360303,0.00003021527,0.00009133608,0.8320848,0.001159366,0.02143809,0.00152712,0.1248071],"study_design_scores_gemma":[0.001261807,0.001350881,0.001333549,0.00005822061,0.0002086645,0.0000157205,0.00002285433,0.9684665,0.001262006,0.02533595,0.0006499648,0.00003375858],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3803948,0.001646317,0.6091135,0.001162848,0.0002109677,0.0009472271,0.000659614,0.0006770433,0.005187756],"genre_scores_gemma":[0.8964522,0.000354529,0.09997895,0.000246422,0.00005422725,0.0008416707,0.0003955147,0.00008474496,0.001591868],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0266127,"threshold_uncertainty_score":0.1407431,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02241980622583802,"score_gpt":0.2689997521670136,"score_spread":0.2465799459411756,"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."}}