Indoor air conditioning method and system for tumor patients

Abstract

The invention provides an indoor air conditioning method and system for tumor patients. The indoor air conditioning method comprises the steps that environment parameter values in a room and physiological parameter values of at least one tumor patient in the room are obtained; based on the environmental parameter values and the physiological parameter values, high-order features are determined, and the high-order features are used for representing the influence relation between the different parameter values; the environment parameter values, the physiological parameter values and the high-order features are input into a regression tree model, target environment parameter values output by the regression tree model are obtained, and the regression tree model is obtained through training based on sample environment parameter values in a plurality of sample rooms and sample physiological parameter values of a plurality of sample tumor patients in the sample rooms; the sample environment parameter value is a parameter value of which the comfort level of the sample tumor patient is greater than a preset comfort level; and based on the target environment parameter value, air in the room is adjusted. The air in the room where the tumor patient is located can be adjusted in a targeted and personalized mode.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, and in particular to an indoor air conditioning method and system for cancer patients. Background Technology

[0002] Cancer patients, due to the disease itself and treatments such as radiotherapy and chemotherapy, generally experience increased resting energy expenditure and metabolic disorders, making them significantly more sensitive to indoor thermal environments than the general population. Specifically, inflammatory factors and metabolic reprogramming weaken patients' thermoregulation abilities, making them more prone to discomfort from hot or cold temperatures. Simultaneously, due to reduced physical activity, oral mucosal damage, and intestinal flora imbalance, patients' needs for temperature, humidity, airflow, and cleanliness exhibit highly individualized characteristics. Therefore, regulating and optimizing the air quality in the rooms where cancer patients reside is crucial.

[0003] In existing technologies, indoor air conditioning involves deploying a regionalized sensor network to acquire environmental data such as temperature and humidity. By combining historical environmental data variations within each area with the initial sensor acquisition frequency, the data update frequency for each area is dynamically adjusted, and temperature thresholds are set. This allows for the calculation of upper and lower limits for temperature and humidity in each area based on the maximum and minimum values ​​collected by the sensors. When environmental parameters exceed the dynamic thresholds, automatic adjustment is triggered, thus regulating the air quality in the room.

[0004] However, the above methods, which only use temperature and humidity thresholds for air conditioning, are difficult to adapt to the high sensitivity of cancer patients to temperature and humidity. Therefore, how to conduct targeted and personalized air conditioning in the rooms of cancer patients is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] This invention provides an indoor air conditioning method and system for cancer patients, which solves the shortcomings of existing technologies that cannot provide targeted and personalized air conditioning for the rooms where cancer patients are located, and achieves targeted and personalized air conditioning for the rooms where cancer patients are located.

[0006] This invention provides an indoor air conditioning method for cancer patients, comprising: Obtain environmental parameter values ​​within the room and physiological parameter values ​​of at least one cancer patient residing in the room; Based on the environmental parameter values ​​and the physiological parameter values, higher-order features are determined, which are used to characterize the influence relationship between different parameter values; The environmental parameter values, the physiological parameter values, and the higher-order features are input into the regression tree model to obtain the target environmental parameter values ​​output by the regression tree model. The regression tree model is trained based on the sample environmental parameter values ​​in multiple sample rooms and the sample physiological parameter values ​​of multiple sample tumor patients in the sample rooms. The sample environmental parameter values ​​are the parameter values ​​where the comfort level of the sample tumor patient is greater than the preset comfort level. The air in the room is adjusted based on the target environmental parameter values.

[0007] According to the present invention, an indoor air conditioning method for cancer patients is provided, wherein the environmental parameter values ​​include ambient temperature, ambient humidity and air velocity, and the physiological parameter values ​​include skin temperature, metabolic rate, body temperature, blood flow and C-reactive protein level.

[0008] According to the present invention, an indoor air conditioning method for cancer patients includes obtaining environmental parameter values ​​in the room and physiological parameter values ​​of at least one cancer patient in the room, comprising: Obtain the initial environmental parameter values ​​in the room and the initial physiological parameter values ​​of each of the tumor patients; The missing parameter values ​​in the initial environmental parameter values ​​and the missing parameter values ​​in the initial physiological parameter values ​​are imputed to obtain the imputed environmental parameter values ​​and the imputed physiological parameter values. The interpolated environmental parameter values ​​and the interpolated physiological parameter values ​​are then subjected to standardization, normalization, and Box-Cox transformation in sequence to obtain the environmental parameter values ​​and the physiological parameter values.

[0009] According to the present invention, an indoor air conditioning method for cancer patients includes interpolating missing parameter values ​​in the initial physiological parameter values, comprising: For a target tumor patient with missing parameter values ​​in the initial physiological parameter values, the distance between the target tumor patient and each other tumor patient is determined based on the unmissing initial physiological parameter values ​​of the target tumor patient and the corresponding physiological parameter values ​​of each other tumor patient. The distance is used to characterize the similarity between the target tumor patient and each of the other tumor patients. Sort all distances in ascending order and select the first preset number of similar tumor patients corresponding to the first set number of distances; Based on the distances to the patients with similar tumors and the physiological parameter values ​​of the patients with similar tumors, the missing parameter values ​​in the initial physiological parameter values ​​of the target tumor patient are imputed.

[0010] According to the present invention, an indoor air conditioning method for cancer patients is provided, wherein the higher-order features include a first higher-order feature, a second higher-order feature, and a third higher-order feature; The determination of higher-order features based on the environmental parameter values ​​and the physiological parameter values ​​includes: Multiplying the ambient temperature value and the ambient humidity value yields the first high-order feature, which characterizes the influence relationship between the ambient temperature value and the ambient humidity value. Multiplying the body temperature value and the metabolic rate value yields the second higher-order feature, which characterizes the influence relationship between the body temperature value and the metabolic rate value. The third higher-order feature is obtained by multiplying the air velocity value and the skin temperature value. The third higher-order feature characterizes the influence relationship between the air velocity value and the skin temperature value.

[0011] According to the present invention, an indoor air conditioning method for cancer patients is provided, wherein the regression tree model is trained based on the following method: The environmental parameter values ​​of samples in multiple sample rooms and the physiological parameter values ​​of samples from tumor patients in each sample room are obtained. The environmental parameter values ​​include the environmental temperature value, the environmental humidity value, and the air velocity value. The regression tree model is initialized based on the average value of the sample environmental parameters in the multiple sample rooms to obtain the initial regression tree model; For each of the sample rooms, the environmental parameter values ​​and physiological parameter values ​​collected in the sample room are input into the initial regression tree model to obtain the predicted environmental parameter values ​​in the sample room output by the initial regression tree model. Based on the sample environmental parameter values ​​in the sample room and the predicted environmental parameter values ​​in the sample room obtained in the previous iteration, the residuals corresponding to each sample room are determined. A new regression tree is trained based on the residuals corresponding to each sample room and the sample physiological parameter values ​​of each tumor patient. The new regression tree is weighted and added to the initial regression tree model to obtain the updated regression tree model. Based on the updated regression tree model, the predicted environmental parameter values ​​in each of the sample rooms are re-predicted. The re-predicted environmental parameter values ​​are used as the new predicted environmental parameter values, and the above steps are repeated until the preset conditions are met. The final model is then determined as the regression tree model. The preset conditions include the number of regression trees reaching a third preset number or the convergence error of the model being less than a preset error.

[0012] According to the present invention, an indoor air conditioning method for cancer patients is provided, wherein the target environmental parameter values ​​include target environmental temperature, target environmental humidity, and target air velocity.

[0013] The present invention also provides an indoor air conditioning system for cancer patients, comprising: The acquisition module is used to acquire environmental parameter values ​​in the room and physiological parameter values ​​of at least one cancer patient in the room; The determination module is used to determine higher-order features based on the environmental parameter values ​​and the physiological parameter values, wherein the higher-order features are used to characterize the influence relationship between different parameter values; The input module is used to input the environmental parameter values, the physiological parameter values ​​and the higher-order features into the regression tree model to obtain the target environmental parameter values ​​output by the regression tree model. The regression tree model is trained based on the sample environmental parameter values ​​in multiple sample rooms and the sample physiological parameter values ​​of multiple sample tumor patients in the sample rooms. The sample environmental parameter values ​​are the parameter values ​​of the comfort level of the sample tumor patients being greater than the preset comfort level. The adjustment module is used to adjust the air in the room based on the target environmental parameter values.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the indoor air conditioning method for cancer patients as described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the indoor air conditioning method for cancer patients as described above.

[0016] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the indoor air conditioning method for cancer patients as described above.

[0017] This invention provides an indoor air conditioning method and system for cancer patients. It acquires environmental parameter values ​​and physiological parameter values ​​of at least one cancer patient in the room. Based on these values, higher-order features are determined to characterize the influence relationships between different parameter values. The environmental parameter values, physiological parameter values, and higher-order features are input into a regression tree model to obtain the target environmental parameter value output by the model. This regression tree model is trained using sample environmental parameter values ​​from multiple sample rooms and sample physiological parameter values ​​from multiple sample cancer patients in those rooms. The sample environmental parameter value represents a value where the comfort level of the sample cancer patient is greater than a preset comfort level. Based on this target environmental parameter value, the air in the room is regulated. Because it constructs higher-order features characterizing the nonlinear coupling relationship between physiological and environmental parameters and dynamically predicts the optimal target environmental parameter value for cancer patients using a regression tree model trained on high-comfort samples, it can accurately identify and rapidly respond to metabolic and thermoregulatory dysfunction in cancer patients, achieving targeted and personalized air conditioning in the rooms where cancer patients reside. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an indoor air conditioning method for cancer patients provided in an embodiment of the present invention.

[0020] Figure 2 This is a schematic diagram of the structure of an indoor air conditioning system for cancer patients provided in an embodiment of the present invention.

[0021] Figure 3 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0023] With advancements in medical technology, cancer treatment has entered a phase of personalization and precision. However, the ward environment significantly impacts patient recovery and comfort during cancer treatment. Currently, air conditioning typically involves sensors acquiring room air quality data, setting air parameter thresholds or user-defined parameters, and comparing the acquired data with these thresholds or parameters to achieve air conditioning. However, this method, relying solely on fixed temperature and humidity thresholds for group-based environmental regulation, struggles to meet the specific comfort needs of cancer patients arising from individual pathological differences and the dynamic evolution of treatment stages, resulting in air conditioning results that do not match the patients' actual needs. Therefore, how to achieve targeted and personalized air conditioning in the rooms of cancer patients is a pressing technical problem that needs to be solved.

[0024] In view of the above-mentioned problems, this invention proposes an indoor air conditioning method for cancer patients. In this method, not only are the environmental parameter values ​​in the room considered, but also the physiological parameter values ​​of at least one cancer patient in the room are obtained. The air in the room is conditioned by combining these two sets of data, so that the air in the room where the cancer patient is located can be conditioned in a targeted and personalized manner.

[0025] The embodiments of this invention can be applied to oncology ward scenarios. The executing entity of this method can be a terminal device, computer, server, server cluster, or an electronic device such as an indoor air conditioning device specifically designed for cancer patients. It can also be an indoor air conditioning device for cancer patients installed in the electronic device, which can be implemented through software, hardware, or a combination of both.

[0026] Figure 1 This is a schematic flowchart of an indoor air conditioning method for cancer patients provided in an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes: Step 101: Obtain the environmental parameter values ​​in the room and the physiological parameter values ​​of at least one cancer patient in the room.

[0027] Among them, environmental parameters include ambient temperature, ambient humidity and air velocity, while physiological parameters include skin temperature, metabolic rate, body temperature, blood flow and C-reactive protein level.

[0028] Specifically, ambient temperature, humidity, and airflow velocity values ​​can be collected using temperature sensors, humidity sensors, and airflow velocity sensors deployed within the room. Skin temperature is obtained using a patch-type skin temperature sensor; oxygen consumption and carbon dioxide emissions are measured using indirect calorimetry to calculate metabolic rate; core body temperature is obtained using a tympanic membrane thermometer or electronic capsule device; blood flow is non-invasively collected using an ultrasonic Doppler flowmeter; and C-reactive protein levels are obtained through blood tests. Of course, these parameters can also be obtained through other methods, and this embodiment of the invention does not limit the specific methods used to obtain these parameters.

[0029] Step 102: Based on environmental and physiological parameter values, determine higher-order features, which are used to characterize the influence relationship between different parameter values.

[0030] In this step, considering the potential interactions between various parameters, it is necessary to determine the mutual influence relationships between different parameters. For example, ambient temperature and humidity jointly affect human comfort and heat dissipation efficiency; there is a strong non-linear relationship between metabolic rate and core body temperature; and airflow velocity affects the skin's heat dissipation efficiency.

[0031] Specifically, interactive high-order features are generated by multiplying environmental and physiological parameter values ​​to explicitly capture the nonlinear coupling effect of the physiological-environmental system in cancer patients.

[0032] Step 103: Input the environmental parameter values, physiological parameter values, and higher-order features into the regression tree model to obtain the target environmental parameter values ​​output by the regression tree model. The regression tree model is trained based on the sample environmental parameter values ​​in multiple sample rooms and the sample physiological parameter values ​​of multiple sample tumor patients in the sample rooms. The sample environmental parameter values ​​are the parameter values ​​of the sample tumor patients whose comfort level is greater than the preset comfort level.

[0033] In this step, the target environmental parameter values ​​include the target environmental temperature, the target environmental humidity, and the target air velocity.

[0034] After obtaining the environmental parameter values, physiological parameter values, and higher-order features, these values ​​can be input into the regression tree model. The regression tree model processes this data and outputs the target environmental parameter value, which is the personalized optimal comfort environmental parameter for cancer patients.

[0035] During training, the regression tree model collects environmental parameter values ​​from multiple sample rooms and physiological parameter values ​​from multiple sample tumor patients within those rooms, and then learns to establish a mapping function from "physiological-environmental state" to "optimal environmental parameters." The environmental parameter values ​​represent the parameters that allow a large number of sample tumor patients to achieve a high level of comfort.

[0036] For example, if the ambient temperature of the current ward is 26℃, the ambient humidity is 60%, and the air velocity is 0.2m / s, and the body temperature of the cancer patient in the current ward is 38.2℃, the skin temperature is 37.5℃, the metabolic rate has increased to 1.6 times the baseline value, the blood flow is 4.2L / min, and the C-reactive protein is 45mg / L, after calculating the higher-order features from these values, the data and the calculated higher-order features are input into a pre-trained regression tree model. Thus, the target ambient temperature output by the regression tree model can be obtained as 22.5℃, the target ambient humidity as 50%, and the target air velocity as 0.4m / s.

[0037] Step 104: Adjust the air in the room based on the target environmental parameter values.

[0038] In this step, the target ambient temperature, humidity, and air velocity values ​​output from the regression tree model are converted into control commands. These commands are then used by a proportional-integral-derivative (PID) controller to drive actuators such as air conditioning units, humidifiers, fresh air systems, and variable frequency fans. The controls are adjusted in stages according to preset rates (e.g., temperature change rate ≤ 0.5℃ / min, humidity change rate ≤ 3% / min), while continuously monitoring feedback data from environmental sensors to form a closed-loop control until the error between the actual environmental parameters and the target values ​​is less than a preset threshold. This adjustment method ensures a smooth and abrupt adjustment process, preventing stress responses in cancer patients caused by drastic environmental changes.

[0039] This invention provides an indoor air conditioning method for cancer patients. It acquires environmental parameter values ​​within a room and physiological parameter values ​​of at least one cancer patient in that room. Based on these environmental and physiological parameter values, a higher-order feature is determined. This higher-order feature characterizes the influence relationship between different parameter values. The environmental parameter values, physiological parameter values, and higher-order feature are input into a regression tree model to obtain the target environmental parameter value output by the regression tree model. This regression tree model is trained based on sample environmental parameter values ​​from multiple sample rooms and sample physiological parameter values ​​from multiple sample cancer patients in those rooms. The sample environmental parameter value represents a value where the comfort level of the sample cancer patient is greater than a preset comfort level. Based on the target environmental parameter value, the air in the room is regulated. Because a higher-order feature characterizing the nonlinear coupling relationship between physiological and environmental parameters is constructed, and the regression tree model trained based on high-comfort samples dynamically predicts the optimal target environmental parameter value for cancer patients, it can accurately identify and quickly respond to abnormal metabolic and thermoregulatory functions in cancer patients, achieving targeted and personalized air conditioning in the rooms where cancer patients reside.

[0040] Furthermore, by comprehensively considering and modeling multiple key parameters such as ambient temperature, humidity, air velocity, and the skin temperature, metabolic rate, body temperature, blood flow, and C-reactive protein levels of cancer patients, this invention can precisely regulate various factors of the air environment. Compared to traditional single-dimensional air conditioning methods, it achieves more precise air conditioning. Moreover, this invention not only considers static environmental factors but also achieves dynamic adjustment by monitoring the physiological parameters of cancer patients in real time. The patient's physiological state may change during treatment, and the real-time feedback model can promptly adjust the air environment, ensuring that the patient is always in the most suitable environment, thus contributing to accelerated recovery.

[0041] This invention can not only improve the treatment effect and quality of life of cancer patients and promote the application of precision medicine and personalized treatment, but also be extended to other diseases and clinical scenarios or telemedicine and family health management, becoming an important part of smart healthcare. Therefore, the air conditioning method provided in this invention has a wide range of applications.

[0042] For example, based on the above embodiments, when obtaining environmental parameter values ​​in a room and physiological parameter values ​​of at least one cancer patient in the room, it can be done in the following manner: The initial environmental parameter values ​​in the room and the initial physiological parameter values ​​of each tumor patient are obtained. The missing parameter values ​​in the initial environmental parameter values ​​and the initial physiological parameter values ​​are imputed to obtain the imputed environmental parameter values ​​and the imputed physiological parameter values. The imputed environmental parameter values ​​and the imputed physiological parameter values ​​are then standardized, normalized and transformed by Box-Cox in sequence to obtain the environmental parameter values ​​and physiological parameter values.

[0043] Specifically, after the initial environmental parameter values ​​in the room and the initial physiological parameter values ​​of each tumor patient are collected by the sensors, some parameter values ​​may be missing due to sensor failure, signal transmission interruption or patient activity interference. Therefore, it is necessary to imput these missing values ​​to construct a complete feature vector.

[0044] For example, when imputing missing parameter values ​​in the initial physiological parameter values, it can be done for a target tumor patient whose initial physiological parameter values ​​have missing parameter values. Based on the unmissing initial physiological parameter values ​​of the target tumor patient and the corresponding physiological parameter values ​​of other tumor patients, the distance between the target tumor patient and other tumor patients is determined. The distance is used to characterize the similarity between the target tumor patient and other tumor patients. All distances are sorted in ascending order, and the first preset number of similar tumor patients corresponding to the distances are selected. Based on the distances corresponding to each similar tumor patient and the physiological parameter values ​​of each similar tumor patient, the missing parameter values ​​in the initial physiological parameter values ​​of the target tumor patient are imputed.

[0045] Similarly, when imputing missing parameter values ​​in the initial environmental parameter values, it can be done for a target room with missing parameter values ​​in the initial environmental parameter values. Based on the initial environmental parameter values ​​of the target room that are not missing and the environmental parameter values ​​of other rooms, the distance between the target room and other rooms is determined. The distance is used to characterize the similarity between the target room and other rooms. All distances are sorted in ascending order, and the similar rooms corresponding to the first second preset number of distances are selected. Based on the distances corresponding to each similar room and the environmental parameter values ​​of each similar room, the missing parameter values ​​in the initial environmental parameter values ​​of the target room are imputed.

[0046] Specifically, for target tumor patient i with missing parameter values ​​in the initial physiological parameter values, the missing parameter values ​​can be imputed based on the K-nearest neighbor weighted imputation principle. First, the Euclidean distance between target tumor patient i and each other tumor patient j needs to be calculated based on the following formula (1): (1) in This represents the distance between the target tumor patient i and all other tumor patients j. The smaller the distance, the more similar the target tumor patient i is to the other tumor patients j in terms of physiological state. This represents the k-th non-missing initial physiological parameter value of the target tumor patient i. Let represent the k-th physiological parameter value corresponding to other tumor patients j, and n represent the number of features, where features include physiological parameters.

[0047] After calculating the distance between the target tumor patient i and all other tumor patients j, all calculated distances are sorted in ascending order, and the first set of a first preset number of other tumor patients with the smallest distances are selected as the set of similar tumor patients. The first preset number can be set according to actual conditions or experience; for example, it can be set to 6, meaning selecting 6 patients with the most similar physiological states.

[0048] For the missing g-th physiological parameter of the target tumor patient i, it can be interpolated using the weighted average method according to the following formula (2): (2) in, This represents the missing g-th physiological parameter in target tumor patient i. Let g represent the g-th physiological parameter of patient m with similar tumors. The distance between the target tumor patient i and the similar tumor patient m, where M represents the total number of similar tumor patients, which is also the first preset number.

[0049] When interpolating the missing parameter values ​​in the initial environmental parameter values, a method similar to that used for physiological parameter interpolation is adopted. First, the distance between the target room with missing parameters and each other room is calculated using formula (1). In other words, the physiological parameter values ​​in formula (1) are replaced with environmental parameter values.

[0050] After determining the distance between the target room and all other rooms, select the second preset number of similar rooms that are most similar to the target room, in ascending order of distance.

[0051] Furthermore, the missing parameters of the target room can be imputed using formula (2).

[0052] It should be noted that multiple interpolation or KNN interpolation can also be used to supplement missing parameter values.

[0053] By using the above interpolation method, data from similar tumor patients or similar rooms can be fully utilized to interpolate missing parameters, effectively addressing the problem of missing values ​​caused by sensor failure or signal interruption, ensuring the integrity of the dataset, and improving the accuracy of the target environmental parameters subsequently determined.

[0054] After interpolating the initial environmental and physiological parameter values, to eliminate the influence of different dimensions, the interpolated environmental and physiological parameter values ​​can be standardized. For example, the interpolated environmental temperature value... Skin temperature value and body temperature value Standardization is performed using formula (3) respectively: (3) in, This represents the data after standardization. This represents the data before standardization. Mean values ​​representing characteristics, such as the mean body temperature of all cancer patients in a room, or the mean ambient temperature of all rooms. Standard deviations that represent characteristics, such as the standard deviation of body temperature of all cancer patients in a room, or the standard deviation of ambient temperature in all rooms.

[0055] Of course, other environmental and physiological parameters can also be standardized in the same way.

[0056] For example, the interpolated ambient humidity and interpolated air velocity value For example, in order to scale the data to the range [0,1] and eliminate the influence of extreme values ​​in the data on the results, normalization can also be performed using formula (4): (4) in, This represents the data after normalization. This represents the data before normalization. This represents the minimum value of a characteristic, such as the minimum ambient humidity or minimum air velocity in all rooms. This indicates the maximum value of a characteristic, such as the maximum ambient humidity or the maximum air velocity in all rooms.

[0057] Of course, other environmental and physiological parameters can also be normalized in the same way.

[0058] Furthermore, the metabolic rate value (MR) and blood flow value after intercalation were used as the basis for further analysis. Taking the imputed C-reactive protein level (CRP) as an example, Box-Cox processing can also be performed using formula (5): (5) in, This represents the data after Box-Cox processing. This represents the data before Box-Cox processing. The transformation parameter is obtained by maximizing the log-likelihood function.

[0059] Of course, other environmental and physiological parameters can also be processed using the Box-Cox method described above.

[0060] In this embodiment, by performing standardization, normalization, and Box-Cox transformation on the imputed physiological environment parameter values ​​and the imputed physiological parameter values ​​in sequence, the differences in dimensionality and numerical scale between different parameters can be eliminated, and each feature can be transformed into a unified statistical distribution space, so that the subsequent regression tree model can fairly evaluate the predictive contribution of each parameter.

[0061] For example, based on the above embodiments, the aforementioned higher-order features include a first higher-order feature, a second higher-order feature, and a third higher-order feature. When determining higher-order features based on environmental parameter values ​​and physiological parameter values, the environmental temperature value and the environmental humidity value can be multiplied to obtain the first higher-order feature, which characterizes the influence relationship between the environmental temperature value and the environmental humidity value. The body temperature value and the metabolic rate value can be multiplied to obtain the second higher-order feature, which characterizes the influence relationship between the body temperature value and the metabolic rate value. The air velocity value and the skin temperature value can be multiplied to obtain the third higher-order feature, which characterizes the influence relationship between the air velocity value and the skin temperature value.

[0062] Specifically, since there may be interaction between the various feature parameters, the ambient temperature and humidity values ​​jointly affect the human body's comfort and heat dissipation efficiency, there is a strong nonlinear relationship between the metabolic rate and body temperature, and the air velocity affects the skin's heat dissipation efficiency, the first higher-order feature, the second higher-order feature, and the third higher-order feature can be determined by the following formulas (6)-(8): (6) (7) (8) in, Indicates the first higher-order feature. Indicates the ambient temperature value. Indicates the ambient humidity value. Indicates the second higher-order feature. Indicates body temperature value. This represents the metabolic rate value. Indicates the third higher-order feature. Indicates the air velocity value. This indicates the skin temperature value.

[0063] The first higher-order feature explicitly encodes the synergistic inhibitory effect of hot and humid environments on the body's heat dissipation efficiency; that is, increased humidity nonlinearly weakens the evaporative heat dissipation capacity under high-temperature conditions. The second higher-order feature characterizes the coupling effect between increased metabolic rate and increased body temperature, which increases the net accumulation of heat in the body. The third higher-order feature quantifies the conditional dependence between airflow velocity and skin temperature; that is, the higher the skin temperature, the more significant the heat dissipation efficiency of airflow.

[0064] In the above approach, by determining higher-order features, the potential nonlinear relationships between features can be explored, the nonlinear coupling relationships between different parameters can be explicitly encoded, and the complex interaction effects can be transformed into quantitative indicators that the model can directly utilize. This significantly improves the prediction accuracy and response speed of the regression tree model for the individualized comfort needs of cancer patients, enhances the learning ability of the regression tree model for complex data patterns, and strengthens the expressive power of features.

[0065] For example, based on the above embodiments, the aforementioned regression tree model is trained in the following manner: The process involves acquiring environmental parameter values ​​from multiple sample rooms, as well as physiological parameter values ​​from tumor patients within those rooms. Environmental parameters include temperature, humidity, and airflow velocity. Based on the average of these environmental parameter values, a regression tree model is initialized to obtain an initial regression tree model. For each sample room, the environmental parameter values ​​and physiological parameter values ​​collected within that room are input into the initial regression tree model to obtain the predicted environmental parameter values ​​for that room. Finally, the predicted environmental parameter values ​​are compared with those obtained in the previous iteration. The predicted environmental parameter values ​​within the current room are used to determine the residuals for each sample room. Based on the residuals for each sample room and the physiological parameter values ​​of each sample tumor patient, a new regression tree is trained. The new regression tree is weighted and added to the initial regression tree model to obtain an updated regression tree model. Based on the updated regression tree model, the predicted environmental parameter values ​​within each sample room are re-predicted. The re-predicted environmental parameter values ​​are used as the new predicted environmental parameter values, and the above steps are repeated until the preset conditions are met. The final model is then determined as the regression tree model. The preset conditions include the number of regression trees reaching a third preset number or the model's convergence error being less than a preset error.

[0066] Specifically, environmental parameter values ​​and physiological parameter values ​​of tumor patients in multiple sample rooms were acquired. The environmental parameters included ambient temperature, humidity, and airflow velocity. The physiological parameters included skin temperature, metabolic rate, body temperature, blood flow, and C-reactive protein level. Ambient temperature, humidity, and airflow velocity were set as parameters to ensure that the comfort level of each tumor patient in the sample room exceeded a preset comfort level.

[0067] After obtaining the sample environment parameter values, the average value of these sample environment parameter values ​​can be determined, and the model can be initialized based on the average value.

[0068] For example, the model can be initialized using the mean of the target variable according to formula (9): (9) in, This represents the prediction results of the initial regression tree. This represents the prediction results of the initial regression tree model. This represents the total number of training samples. This represents the value of the target variable in the i-th sample room, which includes the sample ambient temperature, sample ambient humidity, or sample air velocity. The following explanation uses the sample ambient temperature as an example. The model training process for target variables including sample ambient humidity and sample air velocity is similar to that for target variables including sample ambient temperature, and will not be elaborated further here.

[0069] When fitting the regression tree, the higher-order features of the sample can be calculated based on formulas (6)-(8), and the environmental parameter values, physiological parameter values ​​and higher-order features of the sample collected in the sample room can be input into the initial regression tree model obtained after initialization to obtain the predicted environmental parameter values ​​of each sample room output by the initial regression tree model. The predicted environmental parameter values ​​include the predicted environmental temperature value, the predicted environmental humidity value and the predicted air velocity value.

[0070] Furthermore, the residuals corresponding to each sample room can be calculated according to formula (10): (10) in, This represents the residual corresponding to the i-th sample room. This represents the predicted environmental parameter values ​​within the sample room obtained in the previous iteration. This represents the environmental parameter values, physiological parameter values, and high-order features of the sample collected in the i-th sample room.

[0071] After calculating the residuals for each sample room, the calculated residuals can be used to train a new regression tree. Specifically, the squared error can be minimized using the following formula (11): (11) in, This represents the predicted environmental parameters output by the k-th regression tree. This represents the predicted environmental parameters output by the (k-1)th regression tree. This represents the environmental parameter values, physiological parameter values, and high-order features of the sample collected in the i-th sample room.

[0072] Furthermore, the new regression tree obtained through training... The weighted values ​​are added to the initial regression tree model to obtain the updated regression tree model. Based on the updated regression tree model, the predicted environmental parameter values ​​in each sample room are re-predicted according to the following formula (12): (12) in, This represents the predicted environmental parameter values ​​output by the regression tree model after k iterations. This represents the predicted environmental parameter values ​​output by the regression tree model after k-1 iterations. This represents the predicted environmental parameters output by the k-th regression tree. This represents the learning rate, which is usually set to a small value. Although this requires more iterations, it can prevent overfitting. For example, it can be set to 0.01.

[0073] The re-predicted environmental parameter values ​​are used as new environmental parameter values. The residuals corresponding to each sample room are re-determined, and a new regression tree is trained based on the new residuals to update the model. The above steps are repeated until the number of regression trees in the final model reaches the third preset number, or the convergence error of the final model is less than the preset error position. After all iterations are completed, the final regression tree model is the combination of all regression trees, which can be expressed by formula (13): (13) in, This represents the predicted environment parameters output by the final trained regression tree model. This represents the prediction environment parameters output by the initialization regression tree model, and K represents the number of regression trees used during training.

[0074] In the above approach, by enhancing the training of the regression tree model and iteratively optimizing the residuals, the prediction accuracy is gradually improved, giving the regression tree model strong nonlinear fitting capabilities and good interpretability. Furthermore, by determining the optimal predicted environmental parameters for the current physiological state of cancer patients using the regression tree model, a nonlinear mapping relationship of dynamic coupling between physiology and environment can be established. This enables accurate identification and rapid response to metabolic disorders and thermoregulatory dysfunction in cancer patients, thereby outputting personalized, mild, and stable environmental regulation strategies.

[0075] The indoor air conditioning system for cancer patients provided by the present invention will be described below. The indoor air conditioning system for cancer patients described below can be referred to in correspondence with the indoor air conditioning method for cancer patients described above.

[0076] Figure 2 This is a schematic diagram of the structure of an indoor air conditioning system for cancer patients provided in an embodiment of the present invention, as shown below. Figure 2 As shown, the indoor air conditioning system 200 for cancer patients includes: The acquisition module 11 is used to acquire environmental parameter values ​​in the room and physiological parameter values ​​of at least one cancer patient in the room; The determination module 12 is used to determine higher-order features based on the environmental parameter values ​​and the physiological parameter values, wherein the higher-order features are used to characterize the influence relationship between different parameter values; Input module 13 is used to input the environmental parameter values, the physiological parameter values ​​and the higher-order features into the regression tree model to obtain the target environmental parameter values ​​output by the regression tree model. The regression tree model is trained based on the sample environmental parameter values ​​in multiple sample rooms and the sample physiological parameter values ​​of multiple sample tumor patients in the sample rooms. The sample environmental parameter values ​​are the parameter values ​​of the comfort level of the sample tumor patients being greater than the preset comfort level. The adjustment module 14 is used to adjust the air in the room based on the target environmental parameter values.

[0077] In one example embodiment, the environmental parameter values ​​include ambient temperature, ambient humidity, and air velocity, and the physiological parameter values ​​include skin temperature, metabolic rate, body temperature, blood flow, and C-reactive protein level.

[0078] In one example embodiment, the acquisition module 11 is specifically used for: Obtain the initial environmental parameter values ​​in the room and the initial physiological parameter values ​​of each of the tumor patients; The missing parameter values ​​in the initial environmental parameter values ​​and the missing parameter values ​​in the initial physiological parameter values ​​are imputed to obtain the imputed environmental parameter values ​​and the imputed physiological parameter values. The interpolated environmental parameter values ​​and the interpolated physiological parameter values ​​are then subjected to standardization, normalization, and Box-Cox transformation in sequence to obtain the environmental parameter values ​​and the physiological parameter values.

[0079] In one example embodiment, the system further includes an interpolation module, which is specifically used for: For a target tumor patient with missing parameter values ​​in the initial physiological parameter values, the distance between the target tumor patient and each other tumor patient is determined based on the unmissing initial physiological parameter values ​​of the target tumor patient and the corresponding physiological parameter values ​​of each other tumor patient. The distance is used to characterize the similarity between the target tumor patient and each of the other tumor patients. Sort all distances in ascending order and select the first preset number of similar tumor patients corresponding to the first set number of distances; Based on the distances to the patients with similar tumors and the physiological parameter values ​​of the patients with similar tumors, the missing parameter values ​​in the initial physiological parameter values ​​of the target tumor patient are imputed.

[0080] In one example embodiment, the higher-order features include a first higher-order feature, a second higher-order feature, and a third higher-order feature; the determining module 12 is specifically used for: Multiplying the ambient temperature value and the ambient humidity value yields the first high-order feature, which characterizes the influence relationship between the ambient temperature value and the ambient humidity value. Multiplying the body temperature value and the metabolic rate value yields the second higher-order feature, which characterizes the influence relationship between the body temperature value and the metabolic rate value. The third higher-order feature is obtained by multiplying the air velocity value and the skin temperature value. The third higher-order feature characterizes the influence relationship between the air velocity value and the skin temperature value.

[0081] In one example embodiment, the regression tree model is trained in the following manner: The environmental parameter values ​​of samples in multiple sample rooms and the physiological parameter values ​​of samples from tumor patients in each sample room are obtained. The environmental parameter values ​​include the environmental temperature value, the environmental humidity value, and the air velocity value. The regression tree model is initialized based on the average value of the sample environmental parameters in the multiple sample rooms to obtain the initial regression tree model; For each of the sample rooms, the environmental parameter values ​​and physiological parameter values ​​collected in the sample room are input into the initial regression tree model to obtain the predicted environmental parameter values ​​in the sample room output by the initial regression tree model. Based on the sample environmental parameter values ​​in the sample room and the predicted environmental parameter values ​​in the sample room obtained in the previous iteration, the residuals corresponding to each sample room are determined. A new regression tree is trained based on the residuals corresponding to each sample room and the sample physiological parameter values ​​of each tumor patient. The new regression tree is weighted and added to the initial regression tree model to obtain the updated regression tree model. Based on the updated regression tree model, the predicted environmental parameter values ​​in each of the sample rooms are re-predicted. The re-predicted environmental parameter values ​​are used as the new predicted environmental parameter values, and the above steps are repeated until the preset conditions are met. The final model is then determined as the regression tree model. The preset conditions include the number of regression trees reaching a third preset number or the convergence error of the model being less than a preset error.

[0082] In one example embodiment, the target environmental parameter values ​​include target environmental temperature, target environmental humidity, and target air velocity.

[0083] The system of this embodiment can be used in any of the methods in the side embodiment of the indoor air conditioning method for cancer patients. Its specific implementation process and technical effects are similar to those in the side embodiment of the indoor air conditioning method for cancer patients. For details, please refer to the detailed description in the side embodiment of the indoor air conditioning method for cancer patients, which will not be repeated here.

[0084] Figure 3 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 3As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other through the communications bus 340. The processor 310 can call logic instructions in the memory 330 to execute an indoor air conditioning method for cancer patients. The method includes: acquiring environmental parameter values ​​in a room and physiological parameter values ​​of at least one cancer patient in the room; determining higher-order features based on the environmental parameter values ​​and the physiological parameter values, the higher-order features being used to characterize the influence relationship between different parameter values; inputting the environmental parameter values, the physiological parameter values, and the higher-order features into a regression tree model to obtain a target environmental parameter value output by the regression tree model, the regression tree model being trained based on sample environmental parameter values ​​in multiple sample rooms and sample physiological parameter values ​​of multiple sample cancer patients in the sample rooms, the sample environmental parameter value being a parameter value indicating that the comfort level of the sample cancer patient is greater than a preset comfort level; and adjusting the air in the room based on the target environmental parameter value.

[0085] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0086] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the indoor air conditioning method for cancer patients provided by the above methods. The method includes: acquiring environmental parameter values ​​in a room and physiological parameter values ​​of at least one cancer patient in the room; determining higher-order features based on the environmental parameter values ​​and the physiological parameter values, wherein the higher-order features are used to characterize the influence relationship between different parameter values; inputting the environmental parameter values, the physiological parameter values, and the higher-order features into a regression tree model to obtain a target environmental parameter value output by the regression tree model, wherein the regression tree model is trained based on sample environmental parameter values ​​in multiple sample rooms and sample physiological parameter values ​​of multiple sample cancer patients in the sample rooms, wherein the sample environmental parameter value is a parameter value in which the comfort level of the sample cancer patient is greater than a preset comfort level; and adjusting the air in the room based on the target environmental parameter value.

[0087] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements an indoor air conditioning method for cancer patients provided by the methods described above. The method includes: acquiring environmental parameter values ​​in a room and physiological parameter values ​​of at least one cancer patient in the room; determining higher-order features based on the environmental parameter values ​​and the physiological parameter values, the higher-order features being used to characterize the influence relationship between different parameter values; inputting the environmental parameter values, the physiological parameter values, and the higher-order features into a regression tree model to obtain a target environmental parameter value output by the regression tree model, wherein the regression tree model is trained based on sample environmental parameter values ​​in multiple sample rooms and sample physiological parameter values ​​of multiple sample cancer patients in the sample rooms, the sample environmental parameter value being a parameter value indicating that the comfort level of the sample cancer patient is greater than a preset comfort level; and adjusting the air in the room based on the target environmental parameter value.

[0088] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0089] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method of conditioning the indoor air for a tumor patient, characterized by, The method comprises: obtaining an environmental parameter value in a room and a physiological parameter value of at least one tumor patient in the room; determining a high-order feature based on the environmental parameter value and the physiological parameter value, the high-order feature being used to represent an influence relationship between different parameter values; inputting the environmental parameter value, the physiological parameter value and the high-order feature into a regression tree model to obtain a target environmental parameter value output by the regression tree model, the regression tree model being obtained by training based on sample environmental parameter values in a plurality of sample rooms and sample physiological parameter values of a plurality of sample tumor patients in the sample rooms, the sample environmental parameter value being a parameter value of the sample tumor patient whose comfort degree is greater than a preset comfort degree; adjusting air in the room based on the target environmental parameter value.

2. The method for conditioning indoor air for a patient with a tumor according to claim 1, wherein, The environmental parameter value comprises an environmental temperature value, an environmental humidity value and an air flow rate value, and the physiological parameter value comprises a skin temperature value, a metabolic rate value, a body temperature value, a blood flow value and a C-reactive protein level value.

3. The method for conditioning indoor air for a patient with a neoplasm according to claim 2, wherein The method comprises: obtaining an initial environmental parameter value in the room and an initial physiological parameter value of each tumor patient; interpolating missing parameter values in the initial environmental parameter value and the initial physiological parameter value to obtain an interpolated environmental parameter value and an interpolated physiological parameter value; sequentially performing standardization processing, normalization processing and Box-Cox transformation on the interpolated environmental parameter value and the interpolated physiological parameter value to obtain the environmental parameter value and the physiological parameter value.

4. The method for conditioning indoor air for a patient with a neoplasm according to claim 3, wherein The method comprises: for a target tumor patient with missing parameter values in the initial physiological parameter value, determining distances between the target tumor patient and each other tumor patient based on the initial physiological parameter values of the target tumor patient without missing values and the physiological parameter values of each other tumor patient, the distance being used to represent the similarity between the target tumor patient and each other tumor patient; sorting all distances in ascending order and selecting similar tumor patients corresponding to the first preset number of distances; interpolating missing parameter values in the initial physiological parameter value of the target tumor patient based on the distances corresponding to each similar tumor patient and the physiological parameter values of each similar tumor patient.

5. The method for conditioning indoor air for a patient with a neoplasm according to claim 2, wherein The high-order feature comprises a first high-order feature, a second high-order feature and a third high-order feature. The method comprises: multiplying the environmental temperature value and the environmental humidity value to obtain the first high-order feature, the first high-order feature representing an influence relationship between the environmental temperature value and the environmental humidity value; multiplying the body temperature value and the metabolic rate value to obtain the second high-order feature, the second high-order feature representing an influence relationship between the body temperature value and the metabolic rate value; multiplying the air flow rate value and the skin temperature value to obtain the third high-order feature, the third high-order feature representing an influence relationship between the air flow rate value and the skin temperature value.

6. The indoor air conditioning method for cancer patients according to any one of claims 1-5, characterized in that, The regression tree model is obtained based on the following manner: obtaining sample environmental parameter values in a plurality of sample rooms and sample physiological parameter values of sample tumor patients in the sample rooms, the sample environmental parameter values including sample environmental temperature values, sample environmental humidity values and sample air flow rate values; initializing a regression tree model based on average values of the sample environmental parameter values in the plurality of sample rooms to obtain an initial regression tree model; for each of the sample rooms, inputting the sample environmental parameter values in the sample room and the sample physiological parameter values collected in the sample room into the initial regression tree model to obtain predicted environmental parameter values in the sample room output by the initial regression tree model; based on the sample environmental parameter values in the sample room and the predicted environmental parameter values in the sample room obtained in the last iteration, determining a residual error corresponding to each of the sample rooms; training a new regression tree based on the residual error corresponding to each of the sample rooms and the sample physiological parameter values of the sample tumor patients; weighting the new regression tree into the initial regression tree model to obtain an updated regression tree model, re-predicting the predicted environmental parameter values in each of the sample rooms based on the updated regression tree model, taking the re-predicted predicted environmental parameter values as new predicted environmental parameter values, and repeatedly executing the above steps until a preset condition is reached, determining the finally obtained model as the regression tree model, and the preset condition including that the number of regression trees reaches a third preset number or the convergence error of the model is less than a preset error.

7. The method of claim 1-5, wherein the method is for a patient with a tumor. The target environmental parameter values include target environmental temperature values, target environmental humidity values and target air flow rate values.

8. An indoor air conditioning system for a tumor patient, characterized by, comprising: an obtaining module, configured to obtain environmental parameter values in a room and physiological parameter values of at least one tumor patient in the room; a determining module, configured to determine a high-order feature based on the environmental parameter values and the physiological parameter values, the high-order feature being used to represent an influence relationship between different parameter values; an inputting module, configured to input the environmental parameter values, the physiological parameter values and the high-order feature into a regression tree model to obtain target environmental parameter values output by the regression tree model, the regression tree model being obtained based on sample environmental parameter values in a plurality of sample rooms and sample physiological parameter values of a plurality of sample tumor patients in the sample rooms, the sample environmental parameter values being parameter values of the sample tumor patients with a comfort level greater than a preset comfort level; an adjusting module, configured to adjust air in the room based on the target environmental parameter values.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the indoor air conditioning method for tumor patients according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the indoor air conditioning method for tumor patients according to any one of claims 1 to 7.

Images