{"id":"52437edd-9f16-480d-894f-5a5c78acce96","arxiv_id":"2607.08260","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Asymptotic small-signal stability of lossless grid-forming inverter networks is equivalent to positive definiteness of a single matrix Ξ combining topology, operating point, and effective q-V droop gains.","lead":"The paper gives a necessary-and-sufficient small-signal stability test for lossless inverter-based grids of any topology: stability equals positive definiteness of one matrix built from topology, operating point, and droop gains. That matrix view also yields local certificates and measures how much safety margin decentralization throws away.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The Reader correctly isolates the strongest claim (Theorems 1–2) and the weakest modeling assumptions (q-V separability + lossless). Those assumptions bound the scope rather than undermine the internal logic: once Conditions 1 hold and the network is lossless, the loop-shifted matrix Ξ fully determines stability, the Schur/cycle reduction is exact under cos>0, and the cone-graph / local certificates follow by dropping positive-semidefinite terms. The IEEE illustrations support the secondary claim that cycle contributions are often small for the limiting mode without overclaiming universality. Absence of code is a reproducibility gap, not a correctness risk for the analytic result. Consequently the ACCEPT / HIGH-confidence verdict on the mathematical content stands; no adjustment is warranted.","tokens_in":18048,"tokens_out":534,"duration_ms":6107,"concrete_test":"Independently re-derive the closed-loop G(s) = s D̃ (sΓ + Ξ)^{-1} Γ and verify that, for the two-inverter homogeneous case of Sec. VI-A, the Sylvester condition det(Υ)>0 recovered from Theorem 2 coincides exactly with the characteristic polynomial of the linearized droop ODEs having all roots in the open left half-plane (excluding the trivial zero). If they match for a dense sample of (kq, cos Δθ°), the reduction is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central N&S claim (Theorem 1: asymptotic stability on D⊥ iff Ξ ≻ 0 on D⊥; Theorem 2 under cos(Δθ°)>0) is internally consistent within the stated model class. The homotopy argument that equates the number of RHP zeros of pdet(Φ(s)) to the number of negative eigenvalues of Ξ is standard and carefully executed: positivity of Γ(s) prevents imaginary-axis crossings, the high-frequency growth condition (Conditions 1.iii) prevents escape to infinity, and the trivial phase mode is correctly projected out via D⊥. The structural device assumption (q-V separability so that the loop shift produces a strictly positive-real D̃) and the lossless network are modeling choices that the paper itself flags and defers; they do not create an internal gap in the theorems as stated. Cycle-correction negligibility is an empirical observation on three IEEE cases, not a claimed universal fact. No hidden algebraic inconsistency or unstated restriction that would falsify the N&S equivalence inside the model class was found.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper derives a necessary and sufficient small-signal stability criterion for lossless inverter-based power grids with heterogeneous q-V droop devices on arbitrary topology. Under Conditions 1 on the modified device transfer functions, Theorem 1 states that the linearized closed-loop system is asymptotically stable on the subspace D⊥ orthogonal to the trivial phase mode if and only if the matrix Ξ (network linear response plus the diagonal voltage-droop shift K_q^{-1}) is positive definite on D⊥. Under the additional assumption cos(Δθ°)>0, Theorem 2 reduces this to positive definiteness of a Schur-complement matrix Ŷ that includes an explicit cycle correction. Omitting the cycle term and applying further diagonal dominance yields decentralized nodal and edge-wise sufficient certificates (Corollaries 3–4), recovered as special cases of earlier work. Graph-theoretic interpretation via a grounded Laplacian of an augmented cone graph is given, and three IEEE test cases illustrate that the cycle contribution is typically small relative to the stability-limiting mode.","tokens_in":18286,"tokens_out":1130,"duration_ms":10710,"significance":"If the modeling assumptions hold, the result closes a long-standing gap: prior necessary-and-sufficient criteria required restrictive device models or topologies, while decentralized certificates were only sufficient. The single-matrix condition Ξ ≻ 0 on D⊥ unifies network topology, operating point, and effective droop gains, and the cone-graph view makes the conservatism of decentralization quantifiable. The homotopy argument of Theorem 1 is carefully written, the cycle analysis is explicit, and the IEEE illustrations give concrete evidence that cycle corrections are often negligible for the limiting mode. These strengths make the framework a useful analytical tool for grid-code design and for assessing how much security margin local criteria discard.","major_comments":[{"comment":"The structural device assumption of Sec. III-C (existence of k_q^i such that the modified transfer function D̃(s) is strictly positive real in C+ and satisfies the high-frequency growth condition) is load-bearing for the entire N&S reduction: without the loop shift that produces Ξ, Theorem 1 does not apply. The paper correctly flags this and recovers droop as the main example, but the manuscript should state more explicitly which common grid-forming controls (e.g., virtual synchronous machine, matching control, or dispatchable virtual oscillator control) admit this separation and under what parameter restrictions, so that the scope of the N&S claim is clear to practitioners.","section":null},{"comment":"Theorem 2 and all subsequent corollaries require cos(θ_i°−θ_j°)>0 on every edge so that Q≻0. Near the loadability boundary this can fail on heavily loaded lines. The paper should either (i) discuss how the criterion degrades when some cosines become non-positive, or (ii) note that the fundamental statement remains Theorem 1 (Ξ≻0 on D⊥), which does not need this sign condition, and that Theorem 2 is only a convenient reduction inside the usual operating regime.","section":null}],"minor_comments":[{"comment":"Fig. 4 caption and the middle-row panels: the notation for the projection x_min^\top Υ_cycle x_min is clear, but the vertical-axis labels mix λ1(Υ_cycle) with the projection; a short legend or consistent scaling would help readers see that the projection is orders of magnitude smaller than the spectral scale.","section":null},{"comment":"Eq. (6) and the definition of R_ii: the sign convention B_ii<0 is stated early, but a brief reminder when R_ii is introduced would avoid momentary confusion for readers who treat the susceptance matrix as a pure Laplacian.","section":null},{"comment":"The two-inverter example (Sec. VI-A) shows that Corollary 3 is tight for homogeneous droops; a one-sentence remark that this tightness is special to the tree case would prevent over-generalization.","section":null},{"comment":"References [10], [11], [14], [15] appear as arXiv preprints with future dates; if they have been published or updated, the bibliographic entries should be refreshed.","section":null},{"comment":"Typographical: 'V oltage' (space after V) appears in Sec. VI-B; 'desynchronization' is spelled correctly in the references but the abstract uses 'synchronization' consistently—minor polish only.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The central theorems are sound within the stated model class; the two major comments are scope clarifications rather than algebraic gaps. The paper is a natural and substantial extension of the authors' earlier sufficient criteria [13]. Fit for a control/power-systems journal is good. I see no reason to reject or demand major rework of the proofs."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The real news is Theorem 1: for this model class, asymptotic stability on the physical subspace is equivalent to positive definiteness of one matrix Ξ that folds topology, operating point, and the effective voltage droops. Theorem 2 then reduces that to a grounded-Laplacian-style matrix plus an explicit cycle correction. That is the gap the literature review correctly flags—prior exact results needed trees, homogeneous devices, or no voltage dynamics—and they close it without inventing new objects.\n\nWhat they do well is keep the bookkeeping honest. The homotopy that equates RHP poles of the closed-loop transfer function to negative eigenvalues of Ξ is standard and carefully written (positivity of Γ blocks imaginary-axis crossings; the high-frequency growth condition blocks escape to infinity; the trivial phase mode is projected out). The cone-graph reading of the cycle-free certificate is clean, and the IEEE 9/30/118 sweeps show that the cycle term is often large in spectrum yet nearly orthogonal to the limiting mode, so the sufficient Υ ≻ 0 bound is almost tight. They also recover the earlier local droop certificate as a special case and quantify how much decentralization costs. Citations look right; self-citation of their own prior sufficient result is used only as the recovered special case.\n\nSoft spots are the ones they already name. Everything rides on the structural q-V separation (so the loop shift produces a strictly positive-real device map) and on a lossless network. If real inverter firmware does not admit that separation, the reduction fails. No code or data is shipped, so the IEEE numbers are not independently checkable. Those are modeling and reproducibility limits, not algebraic holes inside the stated class.\n\nThis is for people writing decentralized certificates or thinking about grid-code language for grid-forming IBRs. A serious referee should see it. I would engage with the matrix condition and the cycle analysis; the lossy extension is the obvious next step.","headline":"Clean N&S small-signal criterion for heterogeneous q-V droop inverters on arbitrary lossless graphs; the math holds and the cycle analysis is useful.","tokens_in":18950,"tokens_out":491,"would_cite":true,"duration_ms":5493,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Asymptotic small-signal stability of lossless inverter grids is exactly positive-definiteness of one matrix built from topology, operating point, and effective droop gains.","keywords":["small-signal stability","inverter-based resources","grid-forming control","necessary and sufficient criteria","algebraic graph theory","cone graph","droop gains","decentralized certificates"],"falsifier":"Construct a lossless multi-inverter network whose devices satisfy the q-V separation yet whose closed-loop Jacobian has a right-half-plane eigenvalue while Ξ remains positive definite on the relevant subspace (or the converse).","tokens_in":18959,"feed_emoji":"⚡","tokens_out":615,"duration_ms":6436,"temperature":0.7,"pith_summary":"As synchronous generators give way to inverter-based resources, operators need clear rules for when a grid remains stable under small disturbances. Most existing certificates only give sufficient conditions and throw away network structure, so they grow conservative under stress. This paper proves that, for lossless grids of arbitrary topology whose inverters admit an effective q-V droop, linear stability is necessary and sufficient for a single matrix Ξ to be positive definite on the physically relevant subspace. That matrix folds together line susceptances, power-flow angles and voltages, and the local droop gains. Interpreting the reduced matrix as the grounded Laplacian of a weighted cone graph immediately yields local node- and edge-wise certificates, and also measures how much security margin is lost by ignoring cycles. On three IEEE test systems the cycle correction is tiny, so the simpler sufficient tests already track the exact boundary closely. The result supplies both an exact central test and a transparent hierarchy of decentralized tests that could inform future grid codes.","feed_headline":"One matrix decides if inverter grids stay stable","feed_subtitle":"Positive-definiteness of topology, load and droop gains is necessary and sufficient for lossless systems","key_machinery":"The matrix Ξ (equivalently its Schur-reduced form Ŷ), which packages network topology, operating-point loadings, and effective droop gains into one quadratic form whose positive-definiteness decides stability.","core_discovery":"For lossless inverter-based power grids whose devices admit the structural q-V separation, the linearized closed-loop system is asymptotically stable if and only if the single matrix Ξ—the network response plus the diagonal voltage-droop shift—is positive definite on the subspace orthogonal to uniform phase shifts. Under the mild loading condition that all cosine angle differences remain positive, this is equivalent to positive-definiteness of an explicitly reduced matrix that is the grounded Laplacian of a weighted cone graph plus a positive-semidefinite cycle correction.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["One matrix is necessary and sufficient for inverter-grid stability","Positive-definiteness of Ξ certifies lossless inverter-grid stability","Cone-graph Laplacian decides asymptotic stability of inverter grids","Graph cycles barely affect exact stability of lossless inverter systems","Topology, droops and load form one matrix that settles grid stability"],"cache_read_input_tokens":128,"weakest_assumption_plain":"Every inverter must admit an effective reactive-voltage droop constant that makes its modified transfer function strictly positive-real, so the voltage feedback can be cleanly shifted onto the network side.","fun_headline_variants_meta":{"raw":{"variants":["One matrix is necessary and sufficient for inverter-grid stability","Positive-definiteness of Ξ certifies lossless inverter-grid stability","Cone-graph Laplacian decides asymptotic stability of inverter grids","Graph cycles barely affect exact stability of lossless inverter systems","Topology, droops and load form one matrix that settles grid stability"]},"model":"grok-4.5","effort":"low","cost_usd":0.005642,"raw_usage":{"total_tokens":1425,"prompt_tokens":727,"num_sources_used":0,"completion_tokens":86,"cost_in_usd_ticks":56420000,"prompt_tokens_details":{"text_tokens":727,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":612,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":727,"tokens_out":86,"duration_ms":5921,"temperature":1.0,"reasoning_tokens":612,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-10T10:26:50.658073+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Construct a lossless multi-inverter network whose devices satisfy the q-V separation yet whose closed-loop Jacobian has a right-half-plane eigenvalue while Ξ remains positive definite on the relevant subspace (or the converse).","supporting_citations":[],"review_version":1}