Collaborative light regulation and control decision-making method and system

Abstract

The invention discloses a collaborative light regulation and control decision method and system, and relates to the technical field of intelligent regulation and control of plant factory light environments, and the system comprises a plant physiology monitoring unit which is provided with a non-invasive microelectrode array and a hyperspectral imager; and the environment sensing unit is used for collecting illumination intensity, spectrum, temperature, humidity and carbon dioxide concentration data. According to the collaborative light regulation and control decision-making method and system, fusion of multi-source heterogeneous data is realized through a pulse time sequence coding mechanism, and the problem that environmental parameters and plant physiological statuses are difficult to collaboratively analyze in the prior art is solved; based on a dynamic optimization architecture of double-alliance Nash equilibrium, autonomous collaborative decision-making of spectrum parameters can be completed in a short time, and compared with a traditional preset light formula, the response speed is improved; decision effectiveness can be synchronously guaranteed on the virtual model and biological entity level, the light suppression misjudgment rate is reduced, and meanwhile the plant light energy utilization rate is stably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control of light environments in plant factories, and specifically to a collaborative light control decision-making method and system. Background Art

[0002] In the field of plant light regulation, existing technologies generally rely on preset light recipes or simple feedback control based on a single environmental parameter. While these methods can provide basic light management, their core flaw is their inability to efficiently and accurately address the highly dynamic and interconnected complex changes during plant growth. Plant physiological states, such as photosynthetic rate, morphological development, and metabolic activity, fluctuate constantly and are influenced in real time by multiple environmental factors, including ambient light intensity, spectrum, temperature and humidity, and carbon dioxide concentration. Preset, static light recipes lack adaptability and are difficult to adapt to the actual needs of plants at specific growth stages or when encountering environmental disturbances. Control logic that relies on single sensor data ignores the synergistic effects of multiple factors, potentially leading to control conflicts and resource waste. More critically, existing systems lack an intelligent decision-making core that can integrate multi-source heterogeneous data in real time, rapidly assess plant status, and dynamically optimize the synergistic effects of multiple light source parameters. This results in delayed system response and crude control, making it impossible to continuously optimize light resource utilization efficiency and plant growth goals amidst dynamic changes. Summary of the Invention

[0003] To achieve the above objectives, the present invention is implemented through the following technical solutions: A collaborative light control decision system, comprising: a plant physiological monitoring unit, which features a non-invasive microelectrode array and a hyperspectral imager; Environmental sensing unit, used to collect light intensity, spectrum, temperature, humidity and carbon dioxide concentration data; a collaborative decision-making processor, an input end of which is connected to the plant physiological monitoring unit and the environmental sensing unit; A multi-band light source array, controlled by the output of a collaborative decision-making processor; The plant physiological monitoring unit captures the leaf intercellular potential fluctuation frequency, vascular bundle ion flow rate and leaf surface fluorescence peak offset.

[0004] After the system is activated, the plant physiological monitoring unit uses a non-invasive microelectrode array to capture the frequency of leaf intercellular potential fluctuations and the rate of vascular ion flow. A hyperspectral imager simultaneously extracts the peak offset of leaf fluorescence. The environmental sensing unit concurrently collects data on light intensity, ambient spectral distribution, temperature, humidity, and carbon dioxide concentration. All raw data is fed into the collaborative decision-making processor, where its built-in pulse timing encoder performs essential fusion: electrophysiological signals are converted into pulse density features, leaf fluorescence data are mapped into pulse waveform envelopes, and environmental parameters are used to generate constrained pulse sequences, forming a unified time-amplitude pulse stream. This pulse stream is fed into the demand quantification module and the constraint perception module, respectively. The former parses the light demand intensity vector for each band, while the latter outputs a temperature-spectral coupling coefficient matrix.

[0005] Preferably, the collaborative decision processor comprises a pulse timing encoder, the input end of which is connected to the plant physiological monitoring unit and the environmental sensing unit; The pulse sequence encoder converts the intercellular potential fluctuation frequency, ion flow rate, leaf fluorescence peak offset and environmental data into a unified time-amplitude pulse sequence.

[0006] Preferably, the collaborative decision processor is further provided with a demand quantification module and a constraint perception module; The demand quantification module analyzes the pulse sequence and outputs the light demand intensity vector of each band; The constraint perception module generates a coupling coefficient matrix between temperature and spectrum.

[0007] Preferably, the collaborative decision processor includes a collaborative decision engine, which takes the light demand intensity vector and the coupling coefficient matrix as input. The collaborative decision engine calculates the light source parameter combination through an asynchronous gradient tracking algorithm, and the light source parameter combination includes light quality ratio, intensity and duty cycle.

[0008] Preferably, the collaborative decision engine defines the combination of light source parameters as a strategy space, constructs a benefit function based on the smoothness of plant electrophysiological signals and the weighted value of environmental energy consumption, and solves the Pareto optimal solution through the Nash equilibrium principle.

[0009] Preferably, the collaborative light regulation decision system also includes: a virtual verification unit, whose input end is connected to the collaborative decision processor; the virtual verification unit has a built-in metabolic flow topology model for deducing the carbon and nitrogen assimilation path; when light inhibition risk or energy deficit is detected, a rollback instruction is sent to the collaborative decision processor.

[0010] Preferably, the biofeedback unit collects the leaf fluorescence dynamics curve after irradiation with a multi-band light source in real time; compares the deviation value of the curve with the pre-stored standard template; and triggers the collaborative decision processor to re-optimize when the deviation value exceeds a threshold.

[0011] Preferably, after receiving the complete input data, the collaborative decision processor outputs the light source control instruction within a period shorter than the plant physiological response time window.

[0012] The collaborative decision-making engine uses the light demand vector and coupling matrix as inputs and performs dynamic optimization based on the Nash equilibrium principle. It defines the red light ratio, blue light intensity, green spectrum width, and far-red light duty cycle as a four-dimensional strategy space. Red and far-red light automatically form a growth-promoting alliance, while blue and green light form an environmental response alliance.

[0013] The profit function is constructed based on a weighted combination of plant physiological homeostasis coefficients and system energy efficiency. The energy consumption weight is automatically increased when the ambient temperature exceeds a critical value. During the optimization process, the inter-alliance game status is monitored in real time. If the growth-promoting alliance's demand value continuously exceeds the constraint limit, a non-cooperative game mode is initiated. A Pareto frontier search is immediately activated upon detecting signs of light saturation. The optimization results are output to the virtual verification unit and the light source execution unit.

[0014] The virtual verification unit uses metabolic flux topology models to deduce physiological responses. This includes: Under high-temperature conditions, it focuses on verifying the reduced-state concentration of photosystem II ubiquinone. Detecting persistent excess concentrations signals a risk of photoinhibition. Under low-light conditions, it focuses on the carbon flux balance of the Calvin cycle, triggering an energy deficit alarm when sucrose synthesis lags. The biofeedback unit simultaneously collects actual leaf fluorescence kinetic curves and compares them with pre-stored healthy templates for waveform similarity. This includes strictly matching morphological features under normal operating conditions, while automatically relaxing deviation tolerances in low-light or high-humidity winter conditions. Detecting distortion in specific wavelength bands triggers equipment fault diagnosis protocols. The results of this dual verification are fed back to the decision-making engine in real time. Metabolic flux deduction anomalies trigger parameter rollbacks, while fluorescence curve deviations beyond limits trigger re-optimization.

[0015] After receiving the final command, the light source execution unit dynamically adjusts the light quality, intensity, and duty cycle of the multi-band light source. It automatically suppresses the blue light output ratio in high-temperature scenarios, increases the red light intensity and extends the cycle when photosynthetic efficiency is insufficient, and prioritizes maintaining the stability of electrophysiological signals under sudden natural light interference.

[0016] Plant physiological data is recollected after each adjustment. If intercellular potential fluctuations fail to converge or fluorescence peak shifts fail to improve, the entire optimization cycle is restarted. During the high humidity of the rainy season, the system increases green light weight to improve leaf transmittance. In response to pest and disease stress, UV pulses are injected to induce resistance, ensuring that the control strategy is both biologically effective and energy-efficient.

[0017] A collaborative light control decision-making method includes the following steps: S1: Acquire plant electrophysiological signals through a microelectrode array and leaf fluorescence data through a hyperspectral imager; simultaneously collect environmental parameters; S2: Encode plant physiological data and environmental data into pulse sequences; S3: Generate light demand intensity vector and environment coupling coefficient matrix according to the pulse sequence; S4: Calculate the light source parameter combination through Nash equilibrium optimization; S5: Lighting control is performed after double confirmation by virtual verification and biofeedback.

[0018] Preferably, the virtual verification simulates the carbon and nitrogen assimilation process through a metabolic flow topology model; and the biofeedback evaluates the regulatory effect through the deviation of the leaf fluorescence kinetic curve.

[0019] The present invention provides a collaborative light control decision-making method and system. It has the following beneficial effects: This collaborative light regulation decision-making method and system realizes the fusion of multi-source heterogeneous data through a pulse timing coding mechanism, solving the problem of difficult collaborative analysis of environmental parameters and plant physiological states in existing technologies; based on the dynamic optimization architecture of dual-alliance Nash equilibrium, it can complete autonomous collaborative decision-making of spectral parameters in a short time, and the response speed is improved compared with traditional preset light recipes; through the dual verification mechanism of metabolic flow deduction and fluorescence feedback, it simultaneously ensures the effectiveness of decisions at the virtual model and biological entity levels for the first time, reduces the misjudgment rate of light inhibition, and steadily improves the utilization rate of plant light energy.

[0020] This collaborative light regulation decision-making method and system enhances the robustness of the system in complex agricultural scenarios. It avoids light damage through a blue light intelligent suppression strategy under high-temperature conditions and maintains photosynthetic efficiency through a red light compensation mechanism in a weak-light environment. It relies on dynamic response to achieve seamless switching of light environments in the event of sudden environmental disturbances, reducing the incidence of crop physiological stress. For extreme conditions such as the high humidity rainy season and low temperature and weak light in winter, innovative green light transmission enhancement and far-red light cycle adjustment technologies have increased the yield of leafy vegetables and shortened the color change period of fruit crops. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the overall framework of the present invention; Figure 2 Schematic diagram of the framework of the collaborative decision-making processor in the present invention; Figure 3 This is a control logic timing diagram of the present invention. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0023] See also Figures 1 to 3 The present invention provides a technical solution: a collaborative light control decision system, comprising: The plant physiological monitoring unit uses a non-invasive microelectrode array to capture the intercellular potential fluctuation frequency and vascular ion flow rate of the target plant leaves in real time, and uses a hyperspectral imager to extract the leaf fluorescence peak offset; the environmental sensing unit synchronously collects current light intensity, environmental spectral distribution, temperature, humidity and carbon dioxide concentration data.

[0024] The collaborative decision-making processor receives raw data from the plant physiological monitoring unit and the environmental sensing unit, and after internal calculations, generates multi-band light source control instructions. These instructions are output to the multi-band light source array, driving it to adjust the light quality ratio, intensity, and duty cycle of the red, blue, green, and far-red LEDs.

[0025] When plants are under high-temperature stress, the collaborative decision-making processor automatically reduces the blue light output ratio to mitigate the risk of photoinhibition. When the peak fluorescence offset on the leaves indicates insufficient light energy utilization, the red light intensity is increased and the photoperiod is extended to improve energy capture. In the event of sudden environmental disturbances, such as sudden changes in natural light due to cloud cover, the system prioritizes maintaining electrophysiological signal stability to maintain plant homeostasis and dynamically compensate for artificial light parameters.

[0026] The entire implementation process forms a closed-loop control of "data acquisition-decision calculation-light source execution", ensuring that multi-source data streams pass through the hardware module in real time.

[0027] The collaborative decision-making processor includes a pulse timing encoder, the input end of which is connected to the plant physiological monitoring unit and the environmental sensing unit; the pulse timing encoder converts the intercellular potential fluctuation frequency, ion flow rate, leaf fluorescence peak offset and environmental data into a unified time-amplitude pulse sequence.

[0028] It should be further explained that, during implementation, the pulse sequence encoder built into the collaborative decision-making processor receives real-time data on leaf intercellular potential fluctuation frequency, vascular ion flow rate, and leaf fluorescence peak offset from the plant physiological monitoring unit, as well as data on light intensity, spectral distribution, temperature, humidity, and carbon dioxide concentration collected by the environmental sensing unit. The encoder first samples the electrophysiological signals and automatically increases the encoding priority of that data stream if it detects a change in the potential fluctuation frequency exceeding 10% within 0.5 seconds. Simultaneously, the leaf fluorescence data captured by the hyperspectral imager is mapped into a pulse waveform whose peak width is negatively correlated with the fluorescence peak offset. Environmental parameters are then converted into constrained pulse envelopes. For example, a high-amplitude pulse burst is generated when the temperature exceeds 30°C, and the pulse duration is extended when the carbon dioxide concentration falls below 400 ppm. All heterogeneous data are unified into a time-amplitude pulse train through a bionic neural transmission mechanism. The potential fluctuation frequency corresponds to the pulse density, the ion flow rate determines the pulse amplitude slope, and the change in ambient temperature controls the pulse envelope decay rate. In rainy weather conditions where the light intensity drops by more than 50%, the encoder automatically compresses the pulse generation cycle to 1 / 5 of the normal state to ensure that the decision response speed is not affected by sudden environmental changes.

[0029] The collaborative decision-making processor also features a demand quantification module and a constraint perception module. The demand quantification module analyzes the pulse sequence and outputs a light demand intensity vector for each wavelength band. The constraint perception module generates a temperature-spectrum coupling coefficient matrix. It should be noted that, in its implementation, after the pulse sequence is input to the demand quantification module, it first identifies pulse density distribution characteristics. These characteristics include: when the intercellular potential fluctuation pulse density exceeds the historical average by 30%, the plant is deemed to be in a state of photosynthetic stress and the output red light demand intensity vector value is increased to 1.5 times the baseline value. If the slope of the vascular ion current pulse amplitude remains below a threshold for five consecutive minutes, the nutrient compensation mechanism is activated, increasing the far-red light demand intensity by 20%. The concurrently operating constraint perception module monitors the environmental pulse envelope characteristics. If it detects that the temperature pulse amplitude consistently exceeds the critical limit and the carbon dioxide pulse duration decreases to 60% of the normal value, it generates a high-temperature-low-carbon coupling coefficient matrix, forcing the blue light output to be capped at 40% of the maximum power. During the high humidity of the rainy season, the module automatically increases the weight of the green light demand vector to improve leaf transmittance. If the peak width of the leaf fluorescence pulse expands to more than twice the standard value, the light capture efficiency is determined to be insufficient, triggering full-spectrum intensity compensation mode. The outputs of the two modules are compared in real time. When the light demand vector and the coupling matrix conflict in the blue light band, the constraint perception module's limiting instructions take precedence to ensure plant safety.

[0030] The collaborative decision processor includes a collaborative decision engine. The collaborative decision engine takes the light demand intensity vector and the coupling coefficient matrix as input. The collaborative decision engine calculates the light source parameter combination through an asynchronous gradient tracking algorithm; the light source parameter combination includes light quality ratio, intensity and duty cycle.

[0031] It should be further explained that, in the specific implementation process, after the collaborative decision-making engine receives the light demand intensity vectors of each band output by the demand quantification module and the environmental coupling coefficient matrix generated by the constraint perception module, it starts the asynchronous gradient tracking algorithm for dynamic optimization, including: first, taking the current light source parameter combination as the starting point, establishing a four-dimensional strategy space in the light quality ratio dimensions of red, blue, green and far-red light; when the light demand vector indicates that the red light demand intensity exceeds 150% of the baseline value, the red light ratio search step is expanded to 3 times the conventional value to accelerate convergence; at the same time, the temperature-spectrum constraint conditions in the environmental coupling coefficient matrix are monitored; if the blue light suppression coefficient in the matrix is ​​marked as activated, the blue light ratio exploration range is forcibly limited to no more than 40% of the total spectrum during the iteration process.

[0032] The algorithm periodically calculates the trend of a profit function, which is a linear superposition of the smoothness of plant electrophysiological signals and ambient energy consumption. If the profit improvement rate falls below 0.5% over three consecutive iterations, an early stop mechanism is triggered, outputting the current Pareto optimal solution as the light source parameter instruction. Under extreme conditions of high summer midday temperatures and increased natural light, the engine automatically increases the ambient energy consumption weight to 0.5, prioritizing reducing the total power of artificial light sources to avoid energy waste. If leaf fluorescence data simultaneously indicates insufficient light energy utilization, a compensation mode is activated. This involves redistributing the spectral ratio while maintaining total power constraints, increasing the red light duty cycle to 120% of nighttime operation mode to compensate for the loss in photosynthetic efficiency.

[0033] The collaborative decision-making engine defines light source parameter combinations as a strategy space, constructs a benefit function based on the smoothness of plant electrophysiological signals and the weighted value of environmental energy consumption, and solves for the Pareto optimal solution through the Nash equilibrium principle. It should be further explained that during the specific implementation process, the collaborative decision-making engine constructs a light source parameter optimization model based on the Nash equilibrium principle: the light quality ratio, intensity, and duty cycle of red, blue, green, and far-red light are defined as the strategy space of the four participants. The red light ratio and far-red light duty cycle form a growth-promoting alliance, while the blue light intensity and green spectrum width form an environmental response alliance.

[0034] The payoff function is composed of the plant physiological steady-state coefficient and the system's energy efficiency. When the environmental coupling coefficient matrix detects a temperature exceeding 32°C, the blue light alliance's energy efficiency weight is automatically increased to 0.6 to mitigate photothermal damage. The optimization process monitors the inter-alliance game state in real time. For example, if the red light alliance's demand vector value exceeds the constraint matrix limit by 20% for three consecutive minutes, a non-cooperative game mode is activated, allowing the red light ratio to exceed the constraint limit while simultaneously reducing its duty cycle weight. If leaf fluorescence pulses show signs of light saturation, a Pareto frontier search mechanism is activated to identify a compromise solution that satisfies the minimum payoff for all alliances.

[0035] Under high humidity conditions during the rainy season, the system automatically assigns additional compensation weight to the green light spectrum width strategy, increasing its duty cycle to 150% of that in a dry environment, alleviating photosynthetic resistance by enhancing leaf transmittance; if a stable solution is not reached after five iterations of the Nash equilibrium, the light source combination plan under similar environmental parameters in the historical optimal solution library will be activated.

[0036] The collaborative light regulation decision-making system also includes a virtual verification unit, whose input is connected to the collaborative decision-making processor; a built-in metabolic flux topology model for deducing carbon and nitrogen assimilation pathways; and a rollback instruction sent to the collaborative decision-making processor when a photoinhibition risk or energy deficit is detected. It should be further explained that in the specific implementation process, after receiving the light source parameter combination output by the collaborative decision-making processor, the virtual verification unit activates the metabolic flux topology model for dynamic deduction, including: first simulating the light energy capture process of the thylakoid membrane electron transport chain. When the red light duty cycle exceeds the historical average of 40% and the ambient temperature is above 28°C, the concentration of reduced ubiquinone in photosystem II is automatically detected. If this concentration exceeds the safety threshold for 30 seconds, a photoinhibition risk is determined and a level 1 warning is generated.

[0037] The Calvin cycle carbon assimilation pathway is simulated simultaneously, monitoring the balance between ribulose-1,5-bisphosphate regeneration rate and sucrose synthesis flux. An energy deficit alarm is triggered when blue light intensity falls below 50% of the demand vector, resulting in insufficient triose phosphate transporter activity. For high-temperature greenhouse summer conditions, the model prioritizes photorespiration branch validation. Specifically, if glycine decarboxylase flux reaches three times the baseline value and the serine reservoir is saturated, a rollback command is immediately sent to the decision processor, forcing a 20% reduction in red light intensity. In scenarios characterized by continuous overcast and rainy weather and lack of natural light, the model additionally activates the crassulacean acid metabolism validation pathway. If the malate decarboxylation rate is detected to be lagging behind the photoperiod phase, the model automatically extends far-red light exposure to 150% of the normal mode to maintain carbon skeleton turnover. All simulation results are periodically refreshed. If three cumulative simulation cycles show an upward trend in the energy deficit index, the current light source solution is terminated and reverted to the previous stable version.

[0038] The biofeedback unit collects the leaf fluorescence kinetic curves after irradiation with a multi-band light source in real time, compares the deviation between the curves and a pre-stored standard template, and triggers the collaborative decision-making processor to re-optimize when the deviation exceeds a threshold. It should be further explained that in the specific implementation process, the biofeedback unit captures the leaf fluorescence kinetic curves after irradiation with a multi-band light source in real time and dynamically compares them with a pre-stored health response template. This includes: when the fluorescence quenching rate of the curve in the 680nm band lags behind the template baseline by 15%, the photochemical reaction efficiency is determined to be insufficient, and a first-level optimization instruction is generated; if the 730nm reoxidation time of photosystem I is more than 20% ahead of the template, an alarm for light energy distribution imbalance is triggered.

[0039] For low-temperature and low-light conditions in greenhouses during winter, the KL divergence deviation threshold is automatically relaxed to 1.8 times the normal value to avoid false alarms due to a lack of natural light. When an abnormal double peak is detected in the 500-600nm band, the equipment fault diagnosis protocol is initiated. That is, if three consecutive samples show the same characteristics, the backup hyperspectral imager is switched to and the healthy template is reconstructed. In the event of an unexpected pest and disease invasion, the unit automatically activates the stress response mode: when the chlorophyll fluorescence parameter Fv / Fm drops below 0.72 and the fluorescence lifetime in the blue-green light band is abnormally prolonged, the red light intensity is immediately increased by 30% and ultraviolet pulses are injected to induce plant resistance. After each adjustment, resampling and verification are performed. If the KL divergence value decreases by less than 5% for two consecutive iterations, a full parameter reset command is sent to the collaborative decision processor to force the start of a new round of Nash equilibrium optimization cycle.

[0040] After receiving complete input data, the collaborative decision-making processor outputs light source control instructions within a period shorter than the plant physiological response time window. It should be further explained that, in the specific implementation process, the collaborative decision-making processor initiates response timing control after receiving complete input data: first, pulse sequence generation and verification are completed. When the electrophysiological signal sampling rate is detected to be less than 1kHz, the processor automatically switches to the backup microelectrode array. Then, the processor performs Nash equilibrium optimization calculations. If the ambient temperature exceeds 32°C, the processor activates the fast convergence mode, which increases the iterative step size of the asynchronous gradient tracking algorithm by 2.5 times the normal value and compresses the spectral ratio search space to a binary strategy domain dominated by red and blue light. Finally, the processor performs virtual verification preloading to predict the probability of light inhibition risk under high temperature conditions before the metabolic flux topology model is activated. When the historical database shows that similar environmental parameters have triggered a level 3 alarm, a preset correction value of 20% attenuation of blue light intensity is injected in advance. For sensor data packet loss caused by strong lightning interference, the system automatically enables the previous valid data filling mechanism: if three consecutive frames of temperature or light data are lost, an alternative value is generated by linear extrapolation based on the change trend in the last 10 seconds, and the pending verification mark is marked after the decision is made. In the low temperature and high humidity scenario in the early morning of winter, the response cycle is allowed to be extended to prioritize spectral accuracy, but redundant time is forcibly retained for the initialization of the biofeedback channel to ensure strict synchronization between leaf fluorescence sampling and control execution. A timestamp log is generated after each response. When the response delay exceeds the preset response time for five consecutive times, the hardware self-test program is automatically triggered and downgraded to safe operation mode.

[0041] A collaborative light control decision-making method includes the following steps: S1: Acquire plant electrophysiological signals through a microelectrode array and leaf fluorescence data through a hyperspectral imager; simultaneously collect environmental parameters; S2: Encode plant physiological data and environmental data into pulse sequences; S3: Generate light demand intensity vector and environment coupling coefficient matrix according to the pulse sequence; S4: Calculate the light source parameter combination through Nash equilibrium optimization; S5: Lighting control is performed after double confirmation by virtual verification and biofeedback.

[0042] It should be further explained that, in the specific implementation process, the collaborative light regulation decision-making method is implemented in the following steps: the intercellular potential fluctuation frequency and vascular bundle ion flow rate of the leaves are captured in real time through a non-invasive microelectrode array, and the leaf fluorescence peak offset is obtained by using a hyperspectral imager, and time synchronization is achieved with five parameters: light intensity, environmental spectrum, temperature, humidity, and carbon dioxide concentration.

[0043] The heterogeneous data are input into a pulse timing encoder for unified conversion, including switching to a backup acquisition channel when the electrophysiological signal sampling rate is lower than 1kHz, generating a pulse waveform envelope for the leaf fluorescence data according to the wavelength-intensity distribution, and automatically enhancing the temperature pulse amplitude weight when the ambient temperature exceeds 28°C.

[0044] The converted pulse sequence is analyzed by the demand quantification module to obtain the red light dominance index and far-red light compensation coefficient. Combined with the spectrum-temperature coupling matrix generated by the constraint perception module, the Nash equilibrium optimization process is initiated. That is: if the red light dominance index is higher than the historical peak by 15% for two consecutive cycles, the red light strategy space search range is expanded to the conventional value of 180%; when the blue light suppression flag in the coupling matrix is ​​activated, the upper limit of the blue light intensity exploration domain is forcibly limited to 30% of the total spectrum.

[0045] The optimization results are fed into the virtual verification unit to execute metabolic flow deduction: photorespiratory flux is prioritized under high temperature and high humidity conditions. When the glycine accumulation rate reaches 2.5 times the baseline value, execution is interrupted and the system reverts to the safe spectral template. Simultaneously, the biofeedback unit compares the KL divergence of the real-time leaf fluorescence curve with the healthy template. Under low winter light conditions, the deviation threshold is relaxed to 1.8 times. However, a failover protocol is immediately initiated if an abnormal double peak in the 500-600nm band is detected. Finally, control instructions drive the multi-band light source array to adjust spectral parameters. After execution, electrophysiological signals are re-collected to verify the results. If the intercellular potential fluctuation frequency does not converge to the safe range, the full process is re-optimized.

[0046] Virtual validation simulated carbon and nitrogen assimilation using a metabolic flux topology model, while biofeedback assessed the control effect through deviations from leaf fluorescence kinetic curves. It should be noted that the virtual validation and biofeedback implementation of the coordinated light control decision-making method proceeded as follows: The metabolic flux topology model dynamically deduced the state of the thylakoid electron transport chain based on the light source parameter combination. When the ambient temperature exceeded 30°C and the red light duty cycle exceeded the historical average of 35%, the concentration of the reduced state of ubiquinone in Photosystem II was automatically monitored. If this concentration exceeded the safety threshold for 15 consecutive seconds, a Level 2 photoinhibition risk was immediately determined, and a spectral rollback command was generated, forcibly reducing the red light intensity to 80% of the baseline value. Simultaneously, Calvin cycle carbon flux balance verification was performed. Under conditions where the blue light intensity was less than 40% of the demand vector, the activity of the triose phosphate transporter was monitored. Specifically, when the sucrose synthesis rate lagged behind the starch accumulation by 25%, a Level 3 energy deficit alarm was triggered, and a far-red light compensation pulse was injected. The biofeedback channel collects the regulated leaf fluorescence dynamics curve in real time. Under weak light conditions in winter, the KL divergence deviation threshold is automatically relaxed to 1.8 times the standard value. However, if double-peak distortion is detected in the 500-600nm band, the fault protocol is immediately activated: if the distortion feature is repeated three times in a row, the backup hyperspectral imager is switched and the healthy template library is reloaded. In the event of sudden pest and disease stress, when the fluorescence parameter Fv / Fm continuously drops below 0.72 and the fluorescence lifetime in the blue-green band is abnormally extended by 20%, the stress response chain is activated, including: increasing the red light intensity by 30% and inserting a 2-minute ultraviolet pulse cycle. The effect is evaluated after each verification. If the electrophysiological signal smoothness index increases by less than 5% for two consecutive cycles, a full parameter reset request is sent to the decision-making layer, triggering a new round of Nash equilibrium optimization cycle.

[0047] A collaborative light control decision-making method includes the following steps: Step S1: Synchronously collect plant physiological data and environmental parameters: Use a non-invasive microelectrode array to capture the leaf intercellular potential fluctuation frequency and vascular bundle ion flow rate, use a hyperspectral imager to extract the leaf fluorescence peak offset, and simultaneously obtain light intensity, environmental spectrum, temperature, humidity and carbon dioxide concentration data.

[0048] Step S2: Execute multi-source data pulse timing encoding: convert electrophysiological signals into pulse density features, map leaf fluorescence data into pulse waveform envelopes, and generate constrained pulse sequences based on environmental parameters to form a unified time-amplitude pulse stream.

[0049] Step S3: Analyze the pulse stream generation decision parameters: the demand quantification module outputs the light demand intensity vector of each band, and the constraint perception module generates a temperature-spectrum coupling coefficient matrix, and automatically enhances the blue light suppression coefficient weight when the temperature exceeds the critical value.

[0050] Step S4: Start Nash equilibrium collaborative optimization: define red light ratio, blue light intensity, green spectrum width, and far-red light duty cycle as a four-dimensional strategy space, divide the growth promotion alliance and the environmental response alliance, and construct a weighted benefit function of plant physiological homeostasis and system energy consumption.

[0051] Step S5: Dynamically perform optimization calculations: monitor the game status between alliances, activate the non-cooperative game mode when the growth-promoting alliance demand value continues to exceed the limit, trigger the Pareto frontier search when signs of light saturation are detected, and output the light source parameter combination.

[0052] Step S6: Virtual verification of metabolic pathways: deduce the state of the thylakoid electron transport chain through the metabolic flow topology model, focus on monitoring the concentration of the reduced state of photosystem II ubiquinone under high temperature conditions, and verify the carbon flux balance of the Calvin cycle in a weak light environment.

[0053] Step S7: Real-time biofeedback verification: Collect the leaf fluorescence dynamics curve after light source irradiation, compare the waveform similarity with the pre-stored healthy template, strictly match the characteristics under normal working conditions, and automatically relax the deviation tolerance under weak light in winter.

[0054] Step S8: Execute dynamic light source control: adjust the multi-band light source parameters according to the final instructions, suppress the blue light output ratio in high temperature scenes, increase the red light intensity and extend the cycle when the photosynthetic efficiency is insufficient, and prioritize maintaining the stability of electrophysiological signals under sudden interference.

[0055] Step S9: Closed-loop processing of verification results: When metabolic flux deduction is abnormal, the parameters are triggered to roll back to the safety template; when the fluorescence curve deviates beyond the limit, the optimization cycle is restarted; when the device failure characteristics continue to appear, the backup acquisition device is switched.

[0056] Step S10: Effect evaluation and re-optimization: Re-collect the intercellular potential fluctuation frequency and leaf fluorescence data. If the physiological indicators do not converge to the safe range or the energy efficiency continues to deteriorate, the full process decision reset is triggered and the historical optimal plan is injected.

[0057] The fusion of multi-source heterogeneous data is achieved through a pulse timing coding mechanism, which solves the problem of the difficulty in collaborative analysis of environmental parameters and plant physiological states in existing technologies; the dynamic optimization architecture based on the dual-alliance Nash equilibrium can complete autonomous collaborative decision-making of spectral parameters in a short period of time, and the response speed is improved compared to traditional preset light recipes; through the dual verification mechanism of metabolic flow deduction and fluorescence feedback, the effectiveness of decisions is simultaneously guaranteed at the virtual model and biological entity levels for the first time, reducing the misjudgment rate of light inhibition and steadily improving the utilization rate of plant light energy.

[0058] The system robustness is enhanced in complex agricultural scenarios. Under high-temperature conditions, light damage is avoided through an intelligent blue light suppression strategy, and photosynthetic efficiency is maintained by using a red light compensation mechanism in a weak light environment. In the event of sudden environmental interference, dynamic response is relied upon to achieve seamless switching of light environments and reduce the incidence of crop physiological stress. For extreme conditions such as the high humidity rainy season and low temperature and weak light in winter, innovative green light transmission enhancement and far-red light cycle adjustment technologies have increased the yield of leafy vegetables and shortened the color change period of fruit crops.

[0059] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0060] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A collaborative light control decision system, characterized in that: include: a plant physiological monitoring unit, which features a non-invasive microelectrode array and a hyperspectral imager; Environmental sensing unit, used to collect light intensity, spectrum, temperature, humidity and carbon dioxide concentration data; a collaborative decision-making processor, an input end of which is connected to the plant physiological monitoring unit and the environmental sensing unit; A multi-band light source array, controlled by the output of a collaborative decision-making processor; The plant physiological monitoring unit captures the leaf intercellular potential fluctuation frequency, vascular bundle ion flow rate and leaf fluorescence peak offset; The collaborative decision-making processor includes a pulse sequence encoder, the input end of which is connected to the plant physiological monitoring unit and the environmental sensing unit; the pulse sequence encoder converts the intercellular potential fluctuation frequency, ion flow rate, leaf fluorescence peak offset and environmental data into a unified time-amplitude pulse sequence; The collaborative decision processor is further provided with a demand quantification module and a constraint perception module. The demand quantification module analyzes the pulse sequence and outputs the light demand intensity vector of each band, and the constraint perception module generates a coupling coefficient matrix between temperature and spectrum.

2. The collaborative light control decision system according to claim 1, characterized in that: The collaborative decision processor includes a collaborative decision engine, which takes the light demand intensity vector and the coupling coefficient matrix as input. The collaborative decision engine calculates the light source parameter combination through an asynchronous gradient tracking algorithm. The light source parameter combination includes light quality ratio, intensity and duty cycle.

3. The collaborative light control decision system according to claim 2, characterized in that: The collaborative decision-making engine defines the combination of light source parameters as a strategy space, constructs a benefit function based on the smoothness of plant electrophysiological signals and the weighted value of environmental energy consumption, and solves the Pareto optimal solution through the Nash equilibrium principle.

4. The collaborative light control decision system according to claim 1, characterized in that: The collaborative light regulation decision system also includes: a virtual verification unit, the input end of which is connected to the collaborative decision processor; the virtual verification unit has a built-in metabolic flow topology model for deducing the carbon and nitrogen assimilation path; when light inhibition risk or energy deficit is detected, a rollback instruction is sent to the collaborative decision processor.

5. The collaborative light control decision system according to claim 1, characterized in that: The collaborative light regulation decision system also includes: a biofeedback unit, which collects the leaf fluorescence dynamics curve after irradiation with a multi-band light source in real time, compares the deviation value of the curve with the pre-stored standard template, and triggers the collaborative decision processor to re-optimize when the deviation value exceeds a threshold.

6. The collaborative light control decision system according to claim 2, characterized in that: After receiving the complete input data, the collaborative decision processor outputs a light source control instruction within a period shorter than the plant physiological response time window.

7. A collaborative light control decision-making method, characterized in that: The steps include: S1: Acquire plant electrophysiological signals through a microelectrode array and leaf fluorescence data through a hyperspectral imager; simultaneously collect environmental parameters; S2: Encode plant physiological data and environmental data into pulse sequences; S3: Generate light demand intensity vector and environment coupling coefficient matrix according to the pulse sequence; S4: Calculate the light source parameter combination through Nash equilibrium optimization; S5: Lighting control is performed after double confirmation by virtual verification and biofeedback.

8. The collaborative light control decision-making method according to claim 7, characterized in that: The virtual verification simulates the carbon and nitrogen assimilation process through a metabolic flow topology model; the biofeedback evaluates the regulatory effect through the deviation of the leaf fluorescence kinetic curve.

Images