{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:V3NRYKQPF7LHU6XI7T65ZEZ2D2","short_pith_number":"pith:V3NRYKQP","schema_version":"1.0","canonical_sha256":"aedb1c2a0f2fd67a7ae8fcfddc933a1eb4507bbce670ec2c601db6b5cbdc5f92","source":{"kind":"arxiv","id":"2406.03995","version":1},"attestation_state":"computed","paper":{"title":"AC4MPC: Actor-Critic Reinforcement Learning for Nonlinear Model Predictive Control","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.SY"],"primary_cat":"eess.SY","authors_text":"Andrea Ghezzi, Jasper Hoffmann, Katrin Baumg\\\"artner, Moritz Diehl, Robert D. McAllister, Rudolf Reiter","submitted_at":"2024-06-06T12:15:51Z","abstract_excerpt":"\\Ac{MPC} and \\ac{RL} are two powerful control strategies with, arguably, complementary advantages. In this work, we show how actor-critic \\ac{RL} techniques can be leveraged to improve the performance of \\ac{MPC}. The \\ac{RL} critic is used as an approximation of the optimal value function, and an actor roll-out provides an initial guess for primal variables of the \\ac{MPC}. A parallel control architecture is proposed where each \\ac{MPC} instance is solved twice for different initial guesses. Besides the actor roll-out initialization, a shifted initialization from the previous solution is used"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2406.03995","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"eess.SY","submitted_at":"2024-06-06T12:15:51Z","cross_cats_sorted":["cs.AI","cs.SY"],"title_canon_sha256":"5fee98f447665083bfc037daa3d219667d35ff4963714f2bb1afa6e2bab13a5d","abstract_canon_sha256":"bb67e51a0c41a17657ad847a05ee6402566b18eeac248cd09296c9a7b43c001b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:28:18.860468Z","signature_b64":"0D7ZSU6Hc3I2LpoxNrSjuszyuWfkQDr6KPCmIwPtlXX9mhytNHYNl7JUZyf4TGcf1o+yCTJL5rfHlTPsu2HiAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aedb1c2a0f2fd67a7ae8fcfddc933a1eb4507bbce670ec2c601db6b5cbdc5f92","last_reissued_at":"2026-07-05T08:28:18.860059Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:28:18.860059Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AC4MPC: Actor-Critic Reinforcement Learning for Nonlinear Model Predictive Control","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.SY"],"primary_cat":"eess.SY","authors_text":"Andrea Ghezzi, Jasper Hoffmann, Katrin Baumg\\\"artner, Moritz Diehl, Robert D. McAllister, Rudolf Reiter","submitted_at":"2024-06-06T12:15:51Z","abstract_excerpt":"\\Ac{MPC} and \\ac{RL} are two powerful control strategies with, arguably, complementary advantages. In this work, we show how actor-critic \\ac{RL} techniques can be leveraged to improve the performance of \\ac{MPC}. The \\ac{RL} critic is used as an approximation of the optimal value function, and an actor roll-out provides an initial guess for primal variables of the \\ac{MPC}. A parallel control architecture is proposed where each \\ac{MPC} instance is solved twice for different initial guesses. Besides the actor roll-out initialization, a shifted initialization from the previous solution is used"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.03995","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2406.03995/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2406.03995","created_at":"2026-07-05T08:28:18.860111+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.03995v1","created_at":"2026-07-05T08:28:18.860111+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.03995","created_at":"2026-07-05T08:28:18.860111+00:00"},{"alias_kind":"pith_short_12","alias_value":"V3NRYKQPF7LH","created_at":"2026-07-05T08:28:18.860111+00:00"},{"alias_kind":"pith_short_16","alias_value":"V3NRYKQPF7LHU6XI","created_at":"2026-07-05T08:28:18.860111+00:00"},{"alias_kind":"pith_short_8","alias_value":"V3NRYKQP","created_at":"2026-07-05T08:28:18.860111+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2504.02710","citing_title":"Rollout Then Optimize: A One-Step Newton Refinement of Learned Policies for Nonlinear Model Predictive Control","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2508.16474","citing_title":"Reinforcement Learning-based Control via Y-wise Affine Neural Networks (YANNs)","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21211","citing_title":"Reinforcement Learning-based Control via Y-wise Affine Neural Networks: Comparative Case Studies for Chemical Processes","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2","json":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2.json","graph_json":"https://pith.science/api/pith-number/V3NRYKQPF7LHU6XI7T65ZEZ2D2/graph.json","events_json":"https://pith.science/api/pith-number/V3NRYKQPF7LHU6XI7T65ZEZ2D2/events.json","paper":"https://pith.science/paper/V3NRYKQP"},"agent_actions":{"view_html":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2","download_json":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2.json","view_paper":"https://pith.science/paper/V3NRYKQP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.03995&json=true","fetch_graph":"https://pith.science/api/pith-number/V3NRYKQPF7LHU6XI7T65ZEZ2D2/graph.json","fetch_events":"https://pith.science/api/pith-number/V3NRYKQPF7LHU6XI7T65ZEZ2D2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2/action/storage_attestation","attest_author":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2/action/author_attestation","sign_citation":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2/action/citation_signature","submit_replication":"https://pith.science/pith/V3NRYKQPF7LHU6XI7T65ZEZ2D2/action/replication_record"}},"created_at":"2026-07-05T08:28:18.860111+00:00","updated_at":"2026-07-05T08:28:18.860111+00:00"}