Millisecond anti-saturation control method and system for solid waste thermal storage material
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XIAN THERMAL POWER RES INST CO LTD
- Filing Date
- 2026-03-30
- Publication Date
- 2026-06-23
Smart Images
Figure CN122260975A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of thermal control, and in particular to a millisecond-level anti-saturation control method and system for solid waste thermal storage materials. Background Technology
[0002] In the field of thermal control of solid waste thermal storage materials, existing technologies are often limited by slow response speed and high energy loss, making it difficult to achieve precise control at the millisecond level, especially in terms of anti-saturation control.
[0003] For example, while existing technologies such as deep learning-based bio-robots have achieved innovations in biological tissue carriers, multimodal sensor arrays, and biomimetic actuation devices, their application in thermal control, particularly millisecond-level anti-saturation control of solid waste thermal storage materials, has yet to provide a concrete solution. Similarly, another existing technology, an intelligent adaptive semiconductor packaging test optimization method, achieves self-consistent optimization of a high-dimensional dynamic test system through quantized topological sensor networks and topological photonic crystal waveguide field reconstruction technology. However, its technical focus is on semiconductor packaging test and does not address the thermal control of solid waste thermal storage materials, especially their millisecond-level anti-saturation control mechanism.
[0004] Therefore, there is a lack of existing technologies for millisecond-level anti-saturation control schemes for solid waste thermal storage materials based on the principle of synaptic plasticity, which would enable rapid response and efficient energy utilization. Summary of the Invention
[0005] This application proposes a millisecond-level anti-saturation control method and system for solid waste thermal storage materials to overcome the deficiencies of the prior art.
[0006] According to a first aspect of the embodiments of this application, a millisecond-level anti-saturation control method for solid waste thermal storage materials is provided, comprising: The thermal field distribution of solid waste thermal storage material is detected in real time by a neuromorphic temperature sensing array, and the temperature gradient in the thermal field is converted into an electrical pulse frequency signal. Based on the synaptic timing-dependent plasticity rule, the power weight allocation of multiple heater units is dynamically adjusted according to the deviation delay between the electrical pulse frequency signal and the preset frequency threshold, wherein the deviation delay is used to reflect the relationship between temperature deviation and time. The electrical pulse frequency signal is monitored in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. An inhibitory postsynaptic potential is generated through the synaptic control engine to adjust the power of the corresponding heater unit in milliseconds. The number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold is accumulated. When the number of consecutive overshoots exceeds the preset threshold, a biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection.
[0007] In some embodiments, the neuromorphic temperature sensing array employs a shape memory alloy microwire array, and the real-time detection of the thermal field distribution of the solid waste thermal storage material via the neuromorphic temperature sensing array includes: The radial martensitic phase transformation temperature of the shape memory alloy microwire array is set as the critical temperature detection point, and the temperature value of the critical temperature detection point is 500±5℃. Temperature-sensitive regions of solid waste thermal storage materials are identified by measuring the resistance change of the shape memory alloy microfilament array, wherein the response speed of the shape memory alloy microfilament array is in the millisecond range and the frequency output range of the shape memory alloy microfilament array is 0-1kHz.
[0008] In some embodiments, converting the temperature gradient present in the thermal field into an electrical pulse frequency signal includes: The temperature gradient distribution is obtained by monitoring the temperature values at different spatial locations in the thermal field. The temperature gradient is quantized into an electrical pulse frequency signal with a frequency range of 0-1kHz, where the frequency value of the electrical pulse frequency signal is used to directly reflect the magnitude of the temperature gradient change.
[0009] In some implementations, the preset frequency threshold is 800 Hz, and the magnitude of the temperature gradient is proportional to the frequency value of the electrical pulse frequency signal.
[0010] In some implementations, the dynamic adjustment rules for the power weight allocation of the plurality of heater units include attenuation-type S-shaped curves and trapezoidal functions.
[0011] In some implementations, the power weight allocation of multiple heater units is dynamically adjusted using an exponential decay model, the expression of which is shown below: ΔW=A + ·exp(-Δt / τ + )-A - ·exp(Δt / τ - ); Where ΔW represents the change in power weight, Δt represents the deviation delay, and A + A - τ represents the long-term enhancement gain coefficient and the long-term suppression gain coefficient, respectively. + τ - These represent the time constants of the enhancement process and the inhibition process, respectively.
[0012] In some embodiments, the step of generating an inhibitory postsynaptic potential via a synaptic control engine to adjust the power of the corresponding heater unit at the millisecond level includes: Repressive postsynaptic potentials are generated using programmable logic devices; The power reduction of the corresponding heater unit is achieved in milliseconds by using a negative pulse width modulation wave.
[0013] In some implementations, the preset threshold number is 3 times, and when the number of consecutive overshoots exceeds the preset threshold number, a biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection, including: Based on the pseudo neurotransmitter depletion model, when the number of consecutive overshoots exceeds 3, calcium ion channel blockade is activated, forced cooling is performed for 1.5s, energy redirection is performed, and the thermoelectric conversion device is driven to convert the overshoot energy into the operating current of the control circuit for closed-loop control of energy self-circulation.
[0014] In some embodiments, the method further includes: Contactless communication is achieved through non-chemical synaptic transmission.
[0015] According to a second aspect of this application, a millisecond-level anti-saturation control system for solid waste thermal storage materials is provided, comprising: The electrical pulse frequency signal conversion module is used to detect the thermal field distribution of solid waste thermal storage material in real time through a neuromorphic temperature sensing array, and convert the temperature gradient in the thermal field into an electrical pulse frequency signal. The synaptic weight learning module is used to dynamically adjust the power weight allocation of multiple heater units based on the synaptic temporal dependence plasticity rule and the deviation delay between the electrical pulse frequency signal and a preset frequency threshold. The deviation delay is used to reflect the relationship between temperature deviation and time. The power regulation module is used to monitor the electrical pulse frequency signal in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. The module then generates an inhibitory postsynaptic potential through the synaptic control engine to adjust the power of the corresponding heater unit in milliseconds. The biological anti-saturation mechanism trigger module is used to accumulate the number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold. When the number of consecutive overshoots exceeds the preset threshold, the biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection.
[0016] The beneficial effects of the millisecond-level anti-saturation control method and system for solid waste thermal storage materials in this application include at least the following: This application embodiment uses a neuromorphic temperature sensing array to detect the thermal field distribution of solid waste thermal storage materials in real time, and converts the temperature gradient in the thermal field into an electrical pulse frequency signal, achieving rapid and accurate perception of temperature field changes. This feature converts continuous physical temperature field information into discrete frequency signals, providing a foundation for subsequent digital processing and millisecond-level control. Simultaneously, this encoding method enhances the signal's anti-interference capability during transmission and processing. By dynamically adjusting the power weight allocation of multiple heater units based on the synaptic temporal dependence plasticity rule and the time delay of the deviation between the electrical pulse frequency signal and a preset frequency threshold, a learning and adaptive capability similar to a biological nervous system is obtained. Furthermore, it can intelligently allocate energy according to the temporal characteristics of temperature deviation, rather than performing fixed proportional adjustments, thereby dynamically optimizing the control strategy and improving overall energy efficiency when facing complex and variable thermal environments. By monitoring the electrical pulse frequency signal in real time, when the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. A suppressive postsynaptic potential is generated through the synaptic control engine, and the power of the corresponding heater unit is adjusted at the millisecond level, achieving predictive suppression of temperature overshoot trends. The control action is shifted from traditional post-deviation correction to pre-overshoot prevention, and through a hardware-level rapid response mechanism, temperature fluctuations are greatly suppressed, improving system stability to the millisecond level. By accumulating the number of consecutive overshoots where the electrical pulse frequency signal exceeds the preset frequency threshold, a biological anti-saturation mechanism is triggered when the number of consecutive overshoots exceeds the preset threshold. This mechanism reduces entropy by performing forced cooling and energy redirection, providing deep anti-saturation and energy self-circulation safety guarantees for the system. This prevents the control system from entering a runaway state under extreme or continuous disturbances, and by utilizing surplus thermal energy feedback, dependence on external energy is reduced, achieving entropy reduction and energy efficiency optimization within the system. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating the millisecond-level anti-saturation control method for solid waste thermal storage materials according to an embodiment of this application.
[0018] Figure 2 This is a schematic diagram of the structure of a neuromorphic sensor array according to an embodiment of this application.
[0019] Figure 3 This is a schematic diagram of the synaptic weight learning process according to an embodiment of this application.
[0020] Figure 4 This is a flowchart illustrating the biological antisaturation mechanism in an embodiment of this application.
[0021] Figure 5 This is a schematic diagram of the millisecond-level anti-saturation control system for solid waste thermal storage materials according to an embodiment of this application. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the millisecond-level anti-saturation control method and system for solid waste thermal storage materials will be described clearly and completely below in conjunction with the accompanying drawings of the embodiments of this application. Obviously, the described embodiments are only some embodiments of the embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.
[0023] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed embodiments of the present application, but merely to illustrate selected embodiments of the present application. Other embodiments obtained by those skilled in the art based on the embodiments of the present application without inventive effort are all within the scope of protection of the embodiments of the present application.
[0024] It can be noted that similar reference numerals and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it will not be further defined and explained in subsequent figures according to the embodiments of this application.
[0025] This application discloses a millisecond-level anti-saturation control method and system for solid waste thermal storage materials. This millisecond-level anti-saturation control method for solid waste thermal storage materials is based on a millisecond-level anti-saturation control system for solid waste thermal storage materials. The purpose is to apply biological neural control to the field of thermal control, utilize synaptic plasticity rules to achieve dynamic adjustment of heater power weight, replace traditional signal transmission with electrical coupling, and achieve system entropy reduction through energy self-circulation, thereby breaking through the limitations of traditional algorithms and realizing millisecond-level anti-saturation control of solid waste thermal storage materials.
[0026] See attached document Figure 1 As shown, the millisecond-level anti-saturation control method for a solid waste thermal storage material includes the following steps 110-140.
[0027] Step 110: The thermal field distribution of the solid waste thermal storage material is detected in real time by a neuromorphic temperature sensing array, and the temperature gradient in the thermal field is converted into an electrical pulse frequency signal.
[0028] For example, the neuromorphic temperature sensing array uses an array of shape memory alloy (AMS) microfilaments to simulate neural dendrites, with a resistance abrupt change point (±0.1Ω) corresponding to a critical temperature (500±5°C).
[0029] In some embodiments, the real-time detection of the thermal field distribution of solid waste thermal storage material via a neuromorphic temperature sensing array includes: setting the radial martensitic phase transformation temperature of the shape memory alloy (e.g., Nitinol) microfilament array as a critical temperature detection point, the temperature value of which is 500±5℃; identifying the temperature-sensitive region of the solid waste thermal storage material by measuring the resistance change of the shape memory alloy microfilament array, wherein the response speed of the shape memory alloy microfilament array is on the order of milliseconds, and the frequency output range of the shape memory alloy microfilament array is 0-1kHz.
[0030] In some implementations, the temperature gradient existing in the thermal field is converted into an electrical pulse frequency signal, including: obtaining the temperature gradient distribution by monitoring the temperature values at different spatial locations in the thermal field; quantizing the temperature gradient into an electrical pulse frequency signal with a frequency range of 0-1kHz, wherein the response speed of the electrical pulse frequency signal is ≤0.2ms, which is far superior to that of traditional sensors.
[0031] Among them, the frequency value of the electrical pulse frequency signal is used to directly reflect the change amplitude of the temperature gradient.
[0032] For example, in some implementations, the temperature gradient present in the thermal field is converted into an electrical pulse frequency signal by a constant current source.
[0033] For example, see Appendix Figure 2 As shown, the solid waste thermal storage material substrate 1 serves as the support and heat source for the sensing array, and its surface texture illustrates its porous or composite structure; the shape memory alloy microfilament array 2 is the core sensing unit, composed of multiple orthogonally arranged nickel-titanium alloy microfilaments, whose resistance changes abruptly at 500±5℃ (the critical point of martensitic phase transformation), thus achieving temperature sensing; the microfilament intersection sensitive unit 3 refers to the point where the vertical and horizontal microfilaments in the microfilament array intersect and connect, which is the core sensitive area where the resistance changes with temperature; the electrode contact point 4 is used to realize the metallized electrode pads for connecting the microfilaments to the external circuit; the constant current source 5 provides a stable and precise driving current for the microfilament array; the multiplexer 6 is used to select the massive parallel sensing signals of the array to a single processing channel according to the timing; the analog-to-digital converter 7 converts the analog voltage signal output by the MUX into a digital signal; the pulse frequency modulator 8 converts the input digital signal into an electrical pulse frequency signal (f) of 0-1kHz, completing the neuromorphic coding. Figure 2This paper demonstrates the real-time detection of the thermal field distribution of solid waste thermal storage materials using a neuromorphic temperature sensing array. The method involves using a shape memory alloy microwire array as a thermal sensor, setting the radial martensitic phase transformation temperature of the nickel-titanium nandrolone (NiTiNOL) alloy to 500±5℃. When the ambient temperature is below this temperature, the microwires are in a contracted state, resulting in low contact resistance. When the ambient temperature is above this temperature, the microwires elongate due to restoring tension, significantly increasing the contact resistance. Based on this, the ambient temperature can be reflected in real time by measuring the output resistance of the microwire array.
[0034] Since the response time of a single microfilament is on the order of milliseconds, and millisecond-level temperature response sensitivity is a necessary requirement for intelligent thermal insulation control, thermal sensors composed of a single or a few microfilaments cannot be used for high-precision, fast-response intelligent thermal insulation control. To improve the temperature response speed, this application constructs a microfilament array containing a large number of microfilaments, and through reasonable circuit design, selects "hot spot" microfilament combinations that are sensitive to temperature response to improve the overall temperature response speed. Specifically, a fixed voltage is applied across the microfilament array, the resistance value of each microfilament is recorded, and the combination of microfilaments with the largest resistance change is selected as the sensitive hot spot.
[0035] Therefore, although the response time of most microfilaments in the embodiments of this application is relatively long (milliseconds), as long as a small portion of them responds quickly ("hot spots"), it can represent the temperature response speed of the entire array. Thus, the embodiments of this application can significantly improve the response speed of the thermal sensor while ensuring a certain level of accuracy.
[0036] It is worth noting that the shape memory alloys in this application are preferably materials capable of magnetoelastic isotropic transformation, while other types of materials are difficult to obtain stable unidirectional martensitic phase transformation. Furthermore, such materials are more prone to lattice distortion at low temperatures, leading to the loss of shape memory effect; therefore, this application avoids using overly purified materials.
[0037] Step 120: Based on the Synaptic Timing Dependent Plasticity (STDP) rule, dynamically adjust the power weight allocation of multiple heater units according to the time delay of the deviation between the electrical pulse frequency signal and the preset frequency threshold.
[0038] The deviation delay is used to reflect the relationship between temperature deviation and time.
[0039] For example, the preset frequency threshold is 800Hz. The magnitude of the temperature gradient is directly proportional to the frequency value of the electrical pulse signal.
[0040] For example, the dynamic adjustment rules for the power weight allocation of the multiple heater units include a decaying S-curve and a trapezoidal function. Specifically, for synaptic reward / penalty rules (i.e., dynamic adjustment rules) where the dynamically adjusted power weight exceeds a certain threshold (e.g., 20%), a decaying S-curve is used for adjustment; for synaptic reward / penalty rules where the dynamically adjusted power weight is below a certain threshold (e.g., 10%), a trapezoidal function is used for adjustment.
[0041] In some implementations, the power weight allocation of multiple heater units is dynamically adjusted using an exponential decay model.
[0042] An exemplary expression for this exponential decay model is shown below: ΔW=A + ·exp(-Δt / τ + )-A - ·exp(Δt / τ - ); Where ΔW represents the change in power weight, Δt represents the time delay of the deviation, and A + A - τ represents the long-term enhancement gain coefficient and the long-term suppression gain coefficient, respectively. + τ - These represent the time constants of the enhancement process and the inhibition process, respectively.
[0043] This step in the application enables weight quantization, giving the system a learning function and enhancing its anti-interference capabilities.
[0044] For example, see Appendix Figure 3 As shown, for example, refer to Appendix Figure 3 The diagram illustrates the basic working principle of a millisecond-level synaptic control engine. This embodiment modifies the thermal sensor into a format that allows for feedback adjustment of the controller, termed a bio-inspirational thermal sensor (BTS). A BTS is essentially a neural network with plastic weights, where each neuron corresponds to a synapse, and the connection weights of neurons are determined by the cross-sectional area of the corresponding synapse. Because synapses exist in both excited and inhibited states, their corresponding resistances have two fixed states. When a synapse is in an inhibited state, the corresponding resistance is higher; when a synapse is in an excited state, the corresponding resistance is lower. Since resistance is a fundamental component of a circuit, a resistance change sequence containing weight information can be obtained by adjusting the circuit structure. Specifically, a preset threshold pulse generator 9 generates a reference clock pulse sequence with a frequency corresponding to the target temperature or a preset frequency threshold; and a sensing signal input interface 10 receives signals from... Figure 2The measured temperature pulse frequency signal (f_measure) of the neural biomimetic sensing array; the timing deviation detector (comparator) 11 is used to compare the timing of the reference pulse and the measured pulse, and output the deviation delay signal (Δt) between the two; the preset threshold pulse sequence (f_threshold) 12 is used as an equally spaced rectangular pulse sequence as the system control reference; the measured temperature pulse sequence (f_measure) 13 is used as a pulse sequence generated according to the actual temperature change, and its phase has a dynamic deviation delay (Δt) relative to the sequence (12); the deviation delay signal path 14 is used to transmit the deviation. The vertical connection path of the differential time delay signal (Δt) transmits the upper detection result to the central core processing unit. The STDP weight mapper 15 is the core signal processing unit, whose function is to calculate and output the corresponding power weight change (ΔW) based on the input differential time delay (Δt) and the exponential decay function relationship defined by the Synaptic Timing Dependency Plasticity (STDP) rule. The power weight output interface 16 is used to transmit the ΔW value output by the STDP weight mapper to the power divider of the heater unit, fully considering the interference of mutual traffic generated when multiple synapses are active at the same time in the thermal environment. This application preferably adopts a non-chemical synaptic information transmission method. By using electrical coupling to synchronously drive and read multiple active synapses, the traffic congestion problem can be effectively avoided. In this way, complex control calculations can be completed by simply controlling the IC to close different synapses according to a specific time sequence. Since the BTS is essentially a neural network, the gradient descent method in computer science can be used to optimize and adjust the network weights, that is, to continuously update the connection weight of each synapse according to the output result.
[0045] Preferably, to avoid device failure due to overactivation, the connection strength of each synapse in this embodiment meets a certain target upper limit. Therefore, the control accuracy of the BTS can increase exponentially with the inverse square of the distance between components. Based on this, the BTS can easily achieve stable control of complex systems.
[0046] It should be noted that BTS is only a specific example of the preferred neural network in the embodiments of this application. Other types of neural networks also have the same characteristics (e.g., LTP / LTD), and most other neural networks are also applicable to the solutions in the embodiments of this application.
[0047] Step 130: Monitor the electrical pulse frequency signal in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. An inhibitory postsynaptic potential is generated through the synaptic control engine to adjust the power of the corresponding heater unit in milliseconds.
[0048] In some implementations, the power of the corresponding heater unit is adjusted in milliseconds by generating an inhibitory postsynaptic potential through a synaptic control engine. This includes: generating an inhibitory postsynaptic potential (IPSP) through a programmable logic device (e.g., a field-programmable gate array FPGA); and reducing the power of the corresponding heater unit by using a negative pulse width modulation (PWM) wave in a millisecond time (e.g., 1.2 ms), thereby realizing a millisecond-level synaptic control engine.
[0049] In some implementations, the method further includes achieving contactless communication via non-chemical synaptic transmission. Based on this, by using electrical coupling instead of traditional signal transmission, the system response time is further reduced to 70-80 ms.
[0050] Step 140: Accumulate the number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold. When the number of consecutive overshoots exceeds the preset threshold, trigger the biological anti-saturation mechanism to reduce entropy by performing forced cooling and energy redirection.
[0051] For example, the preset number of times threshold is 3 times.
[0052] In some implementations, when the number of consecutive overshoots exceeds a preset threshold, a biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection. This includes: based on a pseudo-neurotransmitter depletion model, when the number of consecutive overshoots exceeds 3, calcium ion channel blockade is activated, forced cooling is performed for 1.5 seconds, energy redirection is then performed, and a thermoelectric conversion device is driven to convert the overshoot energy (i.e., waste heat) into the operating current of the control circuit for closed-loop control of energy self-circulation. This embodiment is further designed as a pseudo-neurotransmitter depletion model.
[0053] For example, see Appendix Figure 4 As shown, the basic working principle of biological antisaturation mechanisms is illustrated. Figure 4To avoid overshoot due to high integration, this application takes into account the plasticity mechanisms such as PCML and TTL in biological systems, and draws on the solution to the overshoot problem in biological nervous systems, that is, to use frequency domain filtering to reduce the impact of noise, so that the entire system can quickly return to a stable state. Among them, the overshoot signal input interface 17 is used to receive the overshoot judgment pulse signal (f_in) from the power regulation module; the continuous overshoot counter 18 is used to count the continuously occurring overshoot events; the threshold comparator 19 is used to determine whether the count value reaches the preset trigger threshold (such as 3 times); the biological anti-saturation state controller 20 is the core control unit, and its internal state changes sequentially when triggered, simulating the processes of "calcium ion channel blockade", "starting forced cooling" and "enabling energy redirection" in turn; the forced cooling actuator 21 is used to receive the controller's instructions to perform timed forced cooling operations on the solid waste thermal storage material; the thermoelectric conversion device 22 is used to receive the controller's enable signal to convert the overshoot thermal energy (waste heat) in the system into electrical energy (I_out); the system energy circulation node 23 is used to feed back the recovered electrical energy output by the thermoelectric conversion device to the system control circuit, forming a closed-loop energy self-circulation path to achieve system entropy reduction.
[0054] It is important to note that an error limiting mechanism can be introduced into the dynamic weight adjustment process based on synaptic temporal dependence plasticity rules. For example, when the error limit is set to a specific value (e.g., 25%), the output of the reward / penalty rule is constrained within this limit, allowing the system to converge to a new equilibrium state after deviation. This mechanism is achieved by discretizing the search space of the control parameters into several intervals and assigning corresponding action probabilities to each interval. During this process, the transition from the current state to the new equilibrium state involves a certain transient process, and its convergence characteristics are jointly determined by two key parameters: the decay rate and the limit value. The decay rate affects the speed at which the system reaches equilibrium, while the limit value directly affects the final steady-state solution of the system.
[0055] This application embodiment uses a neuromorphic temperature sensing array to detect the thermal field distribution of solid waste thermal storage materials in real time, and converts the temperature gradient in the thermal field into an electrical pulse frequency signal, achieving rapid and accurate perception of temperature field changes. This feature converts continuous physical temperature field information into discrete frequency signals, providing a foundation for subsequent digital processing and millisecond-level control. Simultaneously, this encoding method enhances the signal's anti-interference capability during transmission and processing. By dynamically adjusting the power weight allocation of multiple heater units based on the synaptic temporal dependence plasticity rule and the time delay of the deviation between the electrical pulse frequency signal and a preset frequency threshold, a learning and adaptive capability similar to a biological nervous system is obtained. Furthermore, it can intelligently allocate energy according to the temporal characteristics of temperature deviation, rather than performing fixed proportional adjustments, thereby dynamically optimizing the control strategy and improving overall energy efficiency when facing complex and variable thermal environments. By monitoring the electrical pulse frequency signal in real time, when the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. A suppressive postsynaptic potential is generated through the synaptic control engine, and the power of the corresponding heater unit is adjusted at the millisecond level, achieving predictive suppression of temperature overshoot trends. The control action is shifted from traditional post-deviation correction to pre-overshoot prevention, and through a hardware-level rapid response mechanism, temperature fluctuations are greatly suppressed, improving system stability to the millisecond level. By accumulating the number of consecutive overshoots where the electrical pulse frequency signal exceeds the preset frequency threshold, a biological anti-saturation mechanism is triggered when the number of consecutive overshoots exceeds the preset threshold. This mechanism reduces entropy by performing forced cooling and energy redirection, providing deep anti-saturation and energy self-circulation safety guarantees for the system. This prevents the control system from entering a runaway state under extreme or continuous disturbances, and by utilizing surplus thermal energy feedback, dependence on external energy is reduced, achieving entropy reduction and energy efficiency optimization within the system.
[0056] See attached document Figure 5 As shown, this application also discloses a millisecond-level anti-saturation control system for solid waste thermal storage materials, used to implement the millisecond-level anti-saturation control method for solid waste thermal storage materials as described above. This millisecond-level anti-saturation control system for solid waste thermal storage materials includes: an electrical pulse frequency signal conversion module 510, a synaptic weight learning module 520, a power adjustment module 530, and a biological anti-saturation mechanism triggering module 540.
[0057] For example, the electrical pulse frequency signal conversion module 510 is used to detect the thermal field distribution of solid waste thermal storage material in real time through a neuromorphic temperature sensing array, and convert the temperature gradient existing in the thermal field into an electrical pulse frequency signal.
[0058] For example, the synaptic weight learning module 520 is used to dynamically adjust the power weight allocation of multiple heater units based on the synaptic temporal dependence plasticity rule and the deviation delay between the electrical pulse frequency signal and a preset frequency threshold, wherein the deviation delay is used to reflect the relationship between temperature deviation and time.
[0059] For example, the power adjustment module 530 is used to monitor the electrical pulse frequency signal in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. The synaptic control engine generates an inhibitory postsynaptic potential to adjust the power of the corresponding heater unit in milliseconds.
[0060] For example, the biological anti-saturation mechanism triggering module 540 is used to accumulate the number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold. When the number of consecutive overshoots exceeds the preset number threshold, the biological anti-saturation mechanism is triggered to achieve entropy reduction by performing forced cooling and energy redirection.
[0061] This application embodiment can realize complex temperature control scenarios with multiple inputs and multiple outputs, and has rapid anti-saturation capability. It integrates multiple independent heater units for temperature regulation, and intelligently allocates control to each sub-unit to work together to achieve the lowest overall energy consumption. Alternatively, the sub-units can assist each other, substitute, or switch to cope with extreme working conditions. This application embodiment realizes adaptive adjustment of intermediate layer synaptic weights without the need for external processors or communication modules, giving the system self-learning characteristics and stronger anti-interference and robustness. This application embodiment introduces the solid metal allergy mummification phenomenon into the system, and improves the system stability to the sub-second level through extremely short time delay, high-precision actuators, and intelligent feedback algorithms.
[0062] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of this application, and this application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of this application, and these modifications and improvements are also considered to be within the scope of protection of this application.
Claims
1. A millisecond-level anti-saturation control method for solid waste thermal storage materials, characterized in that, include: The thermal field distribution of solid waste thermal storage material is detected in real time by a neuromorphic temperature sensing array, and the temperature gradient in the thermal field is converted into an electrical pulse frequency signal. Based on the synaptic timing-dependent plasticity rule, the power weight allocation of multiple heater units is dynamically adjusted according to the deviation delay between the electrical pulse frequency signal and the preset frequency threshold, wherein the deviation delay is used to reflect the relationship between temperature deviation and time. The electrical pulse frequency signal is monitored in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. An inhibitory postsynaptic potential is generated through the synaptic control engine to adjust the power of the corresponding heater unit in milliseconds. The number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold is accumulated. When the number of consecutive overshoots exceeds the preset threshold, a biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection.
2. The method according to claim 1, wherein the neuromorphic temperature sensing array employs a shape memory alloy microwire array, characterized in that, The real-time detection of the thermal field distribution of solid waste thermal storage materials via a neuromorphic temperature sensing array includes: The radial martensitic phase transformation temperature of the shape memory alloy microwire array is set as the critical temperature detection point, and the temperature value of the critical temperature detection point is 500±5℃. Temperature-sensitive regions of solid waste thermal storage materials are identified by measuring the resistance change of the shape memory alloy microfilament array, wherein the response speed of the shape memory alloy microfilament array is in the millisecond range and the frequency output range of the shape memory alloy microfilament array is 0-1kHz.
3. The method according to claim 1, characterized in that, The process of converting the temperature gradient existing in the thermal field into an electrical pulse frequency signal includes: The temperature gradient distribution is obtained by monitoring the temperature values at different spatial locations in the thermal field. The temperature gradient is quantized into an electrical pulse frequency signal with a frequency range of 0-1kHz, where the frequency value of the electrical pulse frequency signal is used to directly reflect the magnitude of the temperature gradient change.
4. The method according to claim 3, characterized in that, The preset frequency threshold is 800Hz, and the magnitude of the temperature gradient is proportional to the frequency value of the electrical pulse frequency signal.
5. The method according to claim 1, characterized in that, The dynamic adjustment rules for the power weight allocation of the multiple heater units include attenuation-type S-shaped curves and trapezoidal functions.
6. The method according to claim 1, wherein the dynamic adjustment of the power weight allocation of multiple heater units is calculated using an exponential decay model, characterized in that, The expression for the exponential decay model is shown below: ΔW=A + ·exp(-Δt / τ + )-A - ·exp(Δt / τ - ); Where ΔW represents the change in power weight, Δt represents the deviation delay, and A + A - τ represents the long-term enhancement gain coefficient and the long-term suppression gain coefficient, respectively. + τ - These represent the time constants of the enhancement process and the inhibition process, respectively.
7. The method according to claim 1, characterized in that, The process of generating an inhibitory postsynaptic potential via a synaptic control engine to adjust the power of the corresponding heater unit at the millisecond level includes: Repressive postsynaptic potentials are generated using programmable logic devices; The power reduction of the corresponding heater unit is achieved in milliseconds by using a negative pulse width modulation wave.
8. The method according to claim 1, wherein the preset number threshold is 3 times, characterized in that, When the number of consecutive overshoots exceeds a preset threshold, a biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection, including: Based on the pseudo neurotransmitter depletion model, when the number of consecutive overshoots exceeds 3, calcium ion channel blockade is activated, forced cooling is performed for 1.5s, energy redirection is performed, and the thermoelectric conversion device is driven to convert the overshoot energy into the operating current of the control circuit for closed-loop control of energy self-circulation.
9. The method according to claim 1, characterized in that, The method further includes: Contactless communication is achieved through non-chemical synaptic transmission.
10. A millisecond-level anti-saturation control system for solid waste thermal storage materials, characterized in that, include: The electrical pulse frequency signal conversion module is used to detect the thermal field distribution of solid waste thermal storage material in real time through a neuromorphic temperature sensing array, and convert the temperature gradient in the thermal field into an electrical pulse frequency signal. The synaptic weight learning module is used to dynamically adjust the power weight allocation of multiple heater units based on the synaptic temporal dependence plasticity rule and the deviation delay between the electrical pulse frequency signal and a preset frequency threshold. The deviation delay is used to reflect the relationship between temperature deviation and time. The power regulation module is used to monitor the electrical pulse frequency signal in real time. When the electrical pulse frequency signal exceeds the preset frequency threshold, it is determined to be an overshoot precursor. The module then generates an inhibitory postsynaptic potential through the synaptic control engine to adjust the power of the corresponding heater unit in milliseconds. The biological anti-saturation mechanism trigger module is used to accumulate the number of consecutive overshoots of the electrical pulse frequency signal exceeding a preset frequency threshold. When the number of consecutive overshoots exceeds the preset threshold, the biological anti-saturation mechanism is triggered to reduce entropy by performing forced cooling and energy redirection.