Field-supervised control and adaptive stabilisation of cyber-physical systems

Abstract

An adaptive control system has sensors measuring field parameters of the interaction between an apparatus and its environment. A module determines, from the field parameters, whether the apparatus is

Description

Technical Field

[0001] The present invention relates to supervisory control of physical and cyber-physical systems. More particularly, it relates to mechanisms for governing adaptive operation of such systems based on measured physical stability, particularly within distributed, decentralised, and infrastructure-scale environments. Background of the Invention

[0002] Modern physical and cyber-physical systems increasingly incorporate adaptive control capabilities. Such systems are no longer limited to fixed control laws or predetermined operational schedules, but are capable of modifying operational behaviour in response to sensed conditions, historical data, optimisation objectives, or internal state changes. These capabilities are now widely deployed within electrical power systems, industrial automation, robotics, transportation systems, data-processing infrastructure, and other forms of critical infrastructure.

[0003] Advances in sensing, communication, and computation have enabled high-resolution measurement and rapid information exchange across multiple layers of physical systems. Sensors capable of measuring electrical, thermal, mechanical, kinematic, temporal, and resource-related parameters are deployed at component, subsystem, and system levels. Communication networks provide connectivity between distributed elements, and computational resources capable of executing adaptive control processes are increasingly embedded directly within physical apparatus.

[0004] As a consequence, adaptive behaviour is no longer confined to experimental or offline optimisation contexts, but is deployed directly within operational environments in which physical consequences arise immediately from system actions.

[0005] Earlier generations of monitoring and control systems primarily served observational, diagnostic, or advisory functions. In contrast, contemporary systems are increasingly authorised to directly influence physical actuation. Control actions may affect power flows, mechanical motion, thermal states, resource allocation, or other physically consequential variables. In many cases, such actions occur autonomously and at timescales that preclude meaningful external intervention.

[0006] This evolution exposes a fundamental technical challenge: adaptive behaviour that is effective when considered in isolation may destabilise the physical system in which it operates, particularly where multiple adaptive components interact through shared physical substrates such as electrical networks, mechanical structures, thermal environments, or communication resources.

[0007] Existing approaches frequently rely on centralised supervision, remote monitoring, or human oversight to ensure safe operation. However, in distributed and infrastructure-scale systems, such approaches face inherent limitations. Latency, bandwidth constraints, partial observability, and component failure can render centralised supervision ineffective or untimely. Reliance on external intervention becomes increasingly impractical as systems operate at higher speeds and finer spatial granularity.

[0008] Improvements in communication and telemetry do not, by themselves, resolve the problem of governing physically consequential adaptive behaviour. The availability of detailed measurements does not inherently prevent instability, oscillation, or cascading failure.

[0009] Various techniques exist for constraining system operation, including predefined safety envelopes, rule-based interlocks, fallback controllers, and emergency shutdown mechanisms. While effective for managing known failure modes or static operating limits, such techniques are often ill-suited to systems whose behaviour evolves dynamically over time.

[0010] In particular, existing mechanisms typically assume that adaptive behaviour is permitted by default and intervened upon only when violations occur. Such reactive approaches may be insufficient in environments where instability develops rapidly, propagates across interconnected components, or arises from cumulative interactions that remain individually within nominal bounds.

[0011] Furthermore, many constraint mechanisms operate at the level of individual actions or control signals, without addressing whether the system as a whole remains physically stable or robust to disturbance during adaptive operation.

[0012] In current systems, there is generally no mechanism by which permission for adaptive modification of control behaviour is conditioned on demonstrated maintenance of physical stability. Adaptive capability is typically granted a priori, and safety mechanisms operate as overlays rather than as governing authorities.

[0013] As adaptive behaviour becomes more prevalent and more tightly coupled to physical actuation, this absence of stability-conditioned governance presents increasing risk. Systems may optimise local objectives while degrading global stability, or may continue to adapt even as operating margins erode, leading to loss of control, reduced robustness, or system failure.

[0014] What is lacking is a supervisory paradigm in which adaptive operation is treated as conditional upon maintaining operation within a physically stable regime, rather than as an assumed capability constrained only after the fact.

[0015] These challenges are particularly acute in distributed systems comprising multiple interacting nodes coupled through shared physical or cyber-physical substrates. In such systems, local actions may influence global behaviour through the underlying physical field, even in the absence of explicit coordination or communication.

[0016] Traditional coordination mechanisms based on exchange of control states or policies may be impractical, undesirable, or insufficient to ensure collective stability. At the same time, purely local control strategies may fail to account for emergent interactions mediated by the shared environment.

[0017] There therefore remains a need for supervisory mechanisms that operate locally, scale with system size, and regulate adaptive behaviour based on measurable physical effects rather than internal representations or assumptions.

[0018] Accordingly, there exists a need for systems and methods that supervise adaptive operation of physical and cyber-physical apparatus in a manner grounded in measured physical stability. Such supervision should evaluate whether a system remains within a stable operating region, quantify proximity to operating limits, and govern the extent to which adaptive modification is permitted.

[0019] In particular, there is a need for supervisory control architectures that: condition permission for adaptive modification on maintenance of physical stability; restrict, suspend, or revert adaptive behaviour when stability margins degrade; support graduated operating modes corresponding to degrees of stability; operate locally without reliance on centralised supervision; and scale to distributed and infrastructure-level deployments.

[0020] The present invention addresses these and other deficiencies in the prior art. Summary of the Invention

[0021] The present invention provides a supervisory control system and corresponding methods for governing adaptive operation of physical and cyber-physical apparatus based on measured physical stability.

[0022] In accordance with the invention, one or more operational field parameters characterising interaction between an apparatus and its operating environment are measured. Such field parameters may include, without limitation, electrical, thermal, mechanical, kinematic, temporal, or resource-related variables. Based on the measured field parameters, a stability evaluation module determines whether the apparatus is operating within a defined stable operating region and may compute a quantitative stability margin representing proximity to one or more operating limits.

[0023] An authority management module conditions permission for adaptive modification of control actions or operational parameters on the determined stability. The authority management module assigns an operational authority level as a function of the stability margin, such that the extent of permitted adaptive modification varies in accordance with the degree of physical stability exhibited by the apparatus.

[0024] An enforcement module is operatively coupled to one or more actuators of the apparatus and is configured to restrict, suspend, bound, or revert adaptive modification when the apparatus departs from the stable operating region or when the stability margin falls below a predetermined threshold. In such circumstances, execution of baseline control behaviour may be maintained while adaptive modification is constrained or disabled until stability is re-established.

[0025] In certain embodiments, the supervisory control system operates in a plurality of operating modes corresponding to ranges of the stability margin. Such modes may include, by way of example, observation modes in which adaptive modification is prohibited, limited operation modes in which adaptive modification is constrained, full operation modes in which adaptive modification is permitted, and containment modes in which adaptive modification is disabled and operational authority is reduced to a minimum.

[0026] The stability evaluation may further account for rates of change of one or more field parameters, such that operational authority is restricted pre-emptively prior to violation of an operating limit. Additionally, the stability margin may be adjusted based on the reliability, availability, or consistency of the measured field parameters, thereby reducing permitted authority when confidence in stability assessment degrades.

[0027] In distributed embodiments, the apparatus may comprise a plurality of nodes coupled via a shared physical or cyber-physical substrate. Stability evaluation, authority management, and enforcement may be performed locally at each node without reliance on a central supervisory controller. Coordination between nodes may arise indirectly through the response of the shared substrate to node actions, without exchange of internal control state or adaptive parameters between nodes.

[0028] The supervisory control system and methods described herein are applicable to a wide range of apparatus, including electrical power networks, energy conversion systems, robotic installations, industrial automation systems, transportation systems, and data-processing infrastructure.

[0029] By conditioning adaptive modification on maintenance of physical stability, the invention provides a supervisory control framework in which adaptive operation is governed by measured physical behaviour, thereby improving robustness and reducing the likelihood of instability in complex and distributed systems. Detailed Description of the Invention

[0030] The invention will now be described in detail by way of example embodiments. These embodiments are provided to illustrate the principles of the invention and are not intended to limit the scope of protection, which is defined by the claims.

[0031] The present invention relates to a supervisory control system for governing adaptive operation of physical and cyber-physical apparatus. The supervisory control system operates by conditioning permission for adaptive modification of control behaviour on measured physical stability of the apparatus within its operating environment. Rather than assuming that adaptive capability may be exercised freely subject only to reactive safety intervention, the invention treats adaptive modification as a governed capability whose availability is determined by demonstrated maintenance of stable physical operation. Measurement of Field Parameters

[0032] In an embodiment, the supervisory control system comprises one or more sensors configured to measure operational field parameters that characterise physical interaction between the apparatus and its operating environment. Such parameters may include, without limitation, electrical, thermal, mechanical, kinematic, temporal, power-related, communication integrity, or resource-utilisation variables. The invention is not limited to any particular sensing modality, provided that the measured parameters are indicative of physical stability or proximity to operating limits relevant to the apparatus.

[0033] Measured field parameters are provided to a stability evaluation module, which determines whether the apparatus is operating within a defined stable operating region. The stable operating region corresponds to conditions under which the apparatus is physically feasible, robust to disturbance, and capable of sustained operation without loss of control or degradation of integrity. Operating limits defining the stable operating region may be derived from design specifications, regulatory constraints, empirical characterisation, or operational experience, and may account for interactions between multiple parameters rather than isolated thresholds. Stability Margin Determination

[0034] In some embodiments, the stability evaluation module computes a quantitative stability margin representing proximity to one or more operating limits. The stability margin may be continuous or discretised and may represent not only absolute distance from a limit but also robustness against disturbance or rate of degradation. Operation deep within the stable operating region may correspond to a high stability margin, whereas operation near a boundary of physical feasibility may correspond to a reduced stability margin. Authority Management

[0035] The supervisory control system further comprises an authority management module that assigns an operational authority level as a function of the determined stability or stability margin. In accordance with the invention, permission for adaptive modification of control actions or operational parameters is conditioned on the operational authority level.

[0036] Adaptive modification may include any adjustment of control behaviour that alters how the apparatus responds to sensed conditions, modifies internal control parameters, or changes operational strategies beyond previously validated configurations.

[0037] The operational authority level may limit the magnitude or rate of actuation, restrict the scope of permissible actions, constrain reversibility of actions, or bound the extent or frequency with which adaptive modification is permitted. As physical stability improves and the stability margin increases, broader adaptive modification may be authorised. Conversely, as the stability margin degrades, the authority management module progressively restricts adaptive modification, thereby constraining destabilising behaviour before physical limits are violated. Enforcement of Authority Constraints

[0038] An enforcement module applies the authority limits determined by the authority management module to the apparatus. The enforcement module is operatively coupled to one or more actuators or control interfaces and ensures that adaptive modification remains within the permitted operational authority level.

[0039] When the stability margin falls below a defined threshold, the enforcement module may suspend further adaptive modification, revert adaptive parameters to previously validated values, or prevent execution of unverified control behaviour. Execution of baseline control behaviour may continue during such suspension, enabling the apparatus to maintain safe and functional operation without further adaptation.

[0040] In certain embodiments, re-enabling adaptive modification after a period of restriction requires that stability be maintained for a defined persistence interval. The persistence interval reduces oscillatory enabling and disabling of adaptive behaviour and ensures that recovery of stability reflects sustained rather than transient conditions. Operating Modes

[0041] In some embodiments, the supervisory control system operates in a plurality of operating modes corresponding to ranges of the stability margin. Such modes may include, by way of example: (a) an observation mode in which adaptive modification is prohibited while monitoring and baseline operation continue; (b) a limited operation mode in which adaptive modification is permitted within constrained bounds; (c) a full operation mode in which adaptive modification is permitted within the stable operating region; and (d) a containment mode in which adaptive modification is disabled and operational authority is minimised.

[0042] Transitions between operating modes may be governed by the stability margin and, where applicable, persistence criteria.

[0043] Upon entry into containment mode, the supervisory control system may initiate containment actions such as actuation derating, isolation of a subsystem, reversion to a predefined safe configuration, or initiation of a controlled safe state. Containment does not necessarily imply shutdown. Supervisory monitoring of field parameters may continue during containment, enabling assessment of recovery conditions and determination of whether operational authority may be restored or further escalation is required. Rate-of-Change and Confidence-Based Stability Assessment

[0044] In some embodiments, the stable operating region further specifies permitted rates of change for one or more field parameters. The stability evaluation module may monitor such rates of change and pre-emptively restrict operational authority when excessive rates are detected, even if absolute operating limits have not yet been reached. This enables early intervention prior to manifestation of instability as a limit violation.

[0045] The stability evaluation module may further adjust the stability margin based on reliability, availability, or consistency of the measured field parameters. For example, if sensor data becomes intermittent, noisy, or inconsistent, the effective stability margin may be reduced, thereby causing the authority management module to restrict adaptive modification even when nominal operating values appear acceptable. In this manner, confidence in stability assessment directly influences permitted operational authority. Distributed and Decentralised Embodiments

[0046] In distributed embodiments, the apparatus may comprise multiple nodes coupled via a shared physical or cyber-physical substrate. Each node may include a local instance of the supervisory control system and may perform stability evaluation, authority management, and enforcement locally without reliance on a central supervisory controller.

[0047] Coordination between nodes may arise indirectly through the response of the shared substrate to node actions rather than through exchange of internal control states, adaptive parameters, or policies. Collective behaviour therefore emerges from each node independently maintaining operation within its local stable operating region while interacting through the shared environment. Technical Effects

[0048] By conditioning adaptive modification on maintenance of physical stability, the supervisory control framework described herein restricts operation to physically feasible regions of behaviour. Adaptive modification that would otherwise drive the apparatus toward instability is curtailed before such regimes are entered. As a consequence, adaptive processes operating under this framework tend to avoid destabilising trajectories and operate predominantly within stable regions, thereby improving robustness, reliability, and convergence without requiring modification of the underlying adaptive mechanisms.

[0049] The supervisory control system and methods described herein are applicable to a wide range of physical and cyber-physical apparatus, including electrical power networks, energy conversion systems, robotic installations, industrial automation systems, transportation systems, and data-processing infrastructure. Brief Description of the Drawings

[0050] Figure 1 illustrates an example supervisory control architecture according to an embodiment of the invention, comprising measurement of operational field parameters, evaluation of physical stability, determination of an operational authority level, and enforcement of authority constraints on adaptive modification of control behaviour prior to actuation of a physical or cyber-physical apparatus. Figure 1 illustrates an example supervisory control architecture according to an embodiment of the invention. Operational field parameters arising from interaction between a physical or cyber-physical apparatus and its operating environment are measured by one or more sensing modules. A stability evaluation module determines a stable operating region and computes a stability margin, rate-of-change metrics, and confidence measures based on the measured field parameters. An authority management module assigns an operational authority level as a function of the determined stability. An enforcement module restricts, suspends, bounds, or reverts adaptive modification of control behaviour in accordance with the assigned authority level prior to actuation through one or more control interfaces. The resulting actuation influences the apparatus and its interaction with the operating environment, thereby forming a closed supervisory control loop in which permission for adaptive modification is conditioned on measured physical stability.

Claims

26Claims1. A supervisory control system for a physical or cyber-physical apparatus, comprising:(a) one or more sensors configured to measure operational field parameters characterising interaction between the apparatus and its operating environment;(b) a stability evaluation module configured to determine, from the measured field parameters, whether the apparatus is operating within a defined stable operating region determined by physical operating constraints;(c) an authority management module configured to conditionally grant, restrict, or revoke permission for adaptive modification of control actions or operational parameters as a function of the determined stability; and(d) an enforcement module operatively coupled to one or more actuators of the apparatus and configured to restrict, suspend, bound, or revert adaptive modification while maintaining execution of baseline control behaviour when the apparatus departs from the stable operating region.

2. The supervisory control system of claim 1, wherein the field parameters comprise one or more of electrical, thermal, mechanical, kinematic, temporal, power-related, or resourceutilisation parameters.

3. The supervisory control system of claim 1 or claim 2, wherein the stability evaluation module computes a continuous or discretised stability margin representing proximity to one or more operating limits.

4. The supervisory control system of claim 3, wherein the authority management module assigns an operational authority level as a function of the stability margin.

5. The supervisory control system of claim 4, wherein the authority level limits at least one of magnitude of actuation, rate of change of actuation, scope of permissible actions, reversibility of actions, or extent of adaptive modification.

6. The supervisory control system of any preceding claim, wherein adaptive modification is prohibited by default and enabled only when the stability margin exceeds a defined threshold.18 05 267. The supervisory control system of any preceding claim, wherein the enforcement module suspends adaptive modification until stability is maintained for a defined persistence interval.

8. The supervisory control system of any preceding claim, wherein the system operates in a plurality of operating modes corresponding to ranges of the stability margin.

9. The supervisory control system of claim 8, wherein the modes comprise an observation mode, a limited operation mode, a full operation mode, and a containment mode.

10. The supervisory control system of claim 9, wherein entry into the containment mode causes at least one of actuation derating, subsystem isolation, reversion to a safe configuration, or initiation of a controlled safe state.

11. The supervisory control system of any preceding claim, wherein the stability evaluation module restricts authority based on rates of change of at least one field parameter.

12. The supervisory control system of any preceding claim, wherein the stability evaluation module adjusts the stability margin based on reliability, availability, or consistency of measured field parameters.

13. The supervisory control system of any preceding claim, wherein the apparatus comprises a plurality of distributed nodes and wherein stability evaluation, authority management, and enforcement are performed locally at each node without reliance on a central supervisory controller.

14. The supervisory control system of claim 13, wherein coordination between nodes arises through response of a shared physical or cyber-physical substrate to node actions.

15. The supervisory control system of any preceding claim, wherein the apparatus comprises an electrical power system.

16. The supervisory control system of any preceding claim, wherein the apparatus comprises a robotic or mechanical system.18 05 2617. The supervisory control system of any preceding claim, wherein the apparatus comprises a data-processing or communication system.

18. A method of supervising operation of a physical or cyber-physical apparatus, comprising:(a) measuring operational field parameters;(b) evaluating stability relative to a defined stable operating region;(c) conditioning permission for adaptive modification of control actions on the evaluated stability; and(d) enforcing restriction or suspension of adaptive modification while maintaining baseline control behaviour.

19. The method of claim 18, further comprising restricting adaptive modification based on rates of change of field parameters.

20. The method of claim 18 or claim 19, further comprising adjusting permission for adaptive modification based on confidence in stability evaluation.

21. Anon-transitory computer-readable medium storing instructions which, when executed by one or more processors, cause performance of the method of any of claims 18 to 20.

22. The supervisory control system of any preceding claim, wherein permission for adaptive modification of control actions or operational parameters is determined in accordance with measured physical stability of the apparatus within its operating environment.A